Merge branch 'float32integ' of code.sysbio.ioc.ee:martin/object-segmentation into float32integ

Reviewed-on: #2

ARCHITECTURE.md

@@ -2,11 +2,11 @@
 
 ## Project Overview
 
-A desktop application for detecting organelles and membrane branching structures in microscopy images using YOLOv8s, with comprehensive training, validation, and visualization capabilities.
+A desktop application for detecting and segmenting organelles and membrane branching structures in microscopy images using YOLOv8s-seg, with comprehensive training, validation, and visualization capabilities including pixel-accurate segmentation masks.
 
 ## Technology Stack
 
-- **ML Framework**: Ultralytics YOLOv8 (YOLOv8s.pt model)
+- **ML Framework**: Ultralytics YOLOv8 (YOLOv8s-seg.pt segmentation model)
 - **GUI Framework**: PySide6 (Qt6 for Python)
 - **Visualization**: pyqtgraph
 - **Database**: SQLite3
@@ -110,6 +110,7 @@ erDiagram
         float x_max
         float y_max
         float confidence
+        text segmentation_mask
         datetime detected_at
         json metadata
     }
@@ -122,6 +123,7 @@ erDiagram
         float y_min
         float x_max
         float y_max
+        text segmentation_mask
         string annotator
         datetime created_at
         boolean verified
@@ -139,7 +141,7 @@ Stores information about trained models and their versions.
 | model_name | TEXT | NOT NULL | User-friendly model name |
 | model_version | TEXT | NOT NULL | Version string (e.g., "v1.0") |
 | model_path | TEXT | NOT NULL | Path to model weights file |
-| base_model | TEXT | NOT NULL | Base model used (e.g., "yolov8s.pt") |
+| base_model | TEXT | NOT NULL | Base model used (e.g., "yolov8s-seg.pt") |
 | created_at | TIMESTAMP | DEFAULT CURRENT_TIMESTAMP | Model creation timestamp |
 | training_params | JSON | | Training hyperparameters |
 | metrics | JSON | | Validation metrics (mAP, precision, recall) |
@@ -159,7 +161,7 @@ Stores metadata about microscopy images.
 | checksum | TEXT | | MD5 hash for integrity verification |
 
 #### **detections** table
-Stores object detection results.
+Stores object detection results with optional segmentation masks.
 
 | Column | Type | Constraints | Description |
 |--------|------|-------------|-------------|
@@ -172,11 +174,12 @@ Stores object detection results.
 | x_max | REAL | NOT NULL | Bounding box right coordinate (normalized 0-1) |
 | y_max | REAL | NOT NULL | Bounding box bottom coordinate (normalized 0-1) |
 | confidence | REAL | NOT NULL | Detection confidence score (0-1) |
+| segmentation_mask | TEXT | | JSON array of polygon coordinates [[x1,y1], [x2,y2], ...] (normalized 0-1) |
 | detected_at | TIMESTAMP | DEFAULT CURRENT_TIMESTAMP | When detection was performed |
 | metadata | JSON | | Additional metadata (processing time, etc.) |
 
 #### **annotations** table
-Stores manual annotations for training data (future feature).
+Stores manual annotations for training data with optional segmentation masks (future feature).
 
 | Column | Type | Constraints | Description |
 |--------|------|-------------|-------------|
@@ -187,6 +190,7 @@ Stores manual annotations for training data (future feature).
 | y_min | REAL | NOT NULL | Bounding box top coordinate (normalized) |
 | x_max | REAL | NOT NULL | Bounding box right coordinate (normalized) |
 | y_max | REAL | NOT NULL | Bounding box bottom coordinate (normalized) |
+| segmentation_mask | TEXT | | JSON array of polygon coordinates [[x1,y1], [x2,y2], ...] (normalized 0-1) |
 | annotator | TEXT | | Name of person who created annotation |
 | created_at | TIMESTAMP | DEFAULT CURRENT_TIMESTAMP | Annotation timestamp |
 | verified | BOOLEAN | DEFAULT 0 | Whether annotation is verified |
@@ -245,8 +249,9 @@ graph TB
 ### Key Components
 
 #### 1. **YOLO Wrapper** ([`src/model/yolo_wrapper.py`](src/model/yolo_wrapper.py))
-Encapsulates YOLOv8 operations:
-- Load pre-trained YOLOv8s model
+Encapsulates YOLOv8-seg operations:
+- Load pre-trained YOLOv8s-seg segmentation model
+- Extract pixel-accurate segmentation masks
 - Fine-tune on custom microscopy dataset
 - Export trained models
 - Provide training progress callbacks
@@ -255,10 +260,10 @@ Encapsulates YOLOv8 operations:
 **Key Methods:**
 ```python
 class YOLOWrapper:
-    def __init__(self, model_path: str = "yolov8s.pt")
+    def __init__(self, model_path: str = "yolov8s-seg.pt")
     def train(self, data_yaml: str, epochs: int, callbacks: dict)
     def validate(self, data_yaml: str) -> dict
-    def predict(self, image_path: str, conf: float) -> list
+    def predict(self, image_path: str, conf: float) -> list  # Returns detections with segmentation masks
     def export_model(self, format: str, output_path: str)
 ```
 
@@ -435,7 +440,7 @@ image_repository:
   allowed_extensions: [".jpg", ".jpeg", ".png", ".tif", ".tiff"]
 
 models:
-  default_base_model: "yolov8s.pt"
+  default_base_model: "yolov8s-seg.pt"
   models_directory: "data/models"
 
 training:

BUILD.md

@@ -0,0 +1,178 @@
+# Building and Publishing Guide
+
+This guide explains how to build and publish the microscopy-object-detection package.
+
+## Prerequisites
+
+```bash
+pip install build twine
+```
+
+## Building the Package
+
+### 1. Clean Previous Builds
+
+```bash
+rm -rf build/ dist/ *.egg-info
+```
+
+### 2. Build Distribution Archives
+
+```bash
+python -m build
+```
+
+This will create both wheel (`.whl`) and source distribution (`.tar.gz`) in the `dist/` directory.
+
+### 3. Verify the Build
+
+```bash
+ls dist/
+# Should show:
+# microscopy_object_detection-1.0.0-py3-none-any.whl
+# microscopy_object_detection-1.0.0.tar.gz
+```
+
+## Testing the Package Locally
+
+### Install in Development Mode
+
+```bash
+pip install -e .
+```
+
+### Install from Built Package
+
+```bash
+pip install dist/microscopy_object_detection-1.0.0-py3-none-any.whl
+```
+
+### Test the Installation
+
+```bash
+# Test CLI
+microscopy-detect --version
+
+# Test GUI launcher
+microscopy-detect-gui
+```
+
+## Publishing to PyPI
+
+### 1. Configure PyPI Credentials
+
+Create or update `~/.pypirc`:
+
+```ini
+[pypi]
+username = __token__
+password = pypi-YOUR-API-TOKEN-HERE
+```
+
+### 2. Upload to Test PyPI (Recommended First)
+
+```bash
+python -m twine upload --repository testpypi dist/*
+```
+
+Then test installation:
+
+```bash
+pip install --index-url https://test.pypi.org/simple/ microscopy-object-detection
+```
+
+### 3. Upload to PyPI
+
+```bash
+python -m twine upload dist/*
+```
+
+## Version Management
+
+Update version in multiple files:
+- `setup.py`: Update `version` parameter
+- `pyproject.toml`: Update `version` field
+- `src/__init__.py`: Update `__version__` variable
+
+## Git Tags
+
+After publishing, tag the release:
+
+```bash
+git tag -a v1.0.0 -m "Release version 1.0.0"
+git push origin v1.0.0
+```
+
+## Package Structure
+
+The built package includes:
+- All Python source files in `src/`
+- Configuration files in `config/`
+- Database schema file (`src/database/schema.sql`)
+- Documentation files (README.md, LICENSE, etc.)
+- Entry points for CLI and GUI
+
+## Troubleshooting
+
+### Import Errors
+If you get import errors, ensure:
+- All `__init__.py` files are present
+- Package structure follows the setup configuration
+- Dependencies are listed in `requirements.txt`
+
+### Missing Files
+If files are missing in the built package:
+- Check `MANIFEST.in` includes the required patterns
+- Check `pyproject.toml` package-data configuration
+- Rebuild with `python -m build --no-isolation` for debugging
+
+### Version Conflicts
+If version conflicts occur:
+- Ensure version is consistent across all files
+- Clear build artifacts and rebuild
+- Check for cached installations: `pip list | grep microscopy`
+
+## CI/CD Integration
+
+### GitHub Actions Example
+
+```yaml
+name: Build and Publish
+
+on:
+  release:
+    types: [created]
+
+jobs:
+  deploy:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v2
+    - uses: actions/setup-python@v2
+      with:
+        python-version: '3.8'
+    - name: Install dependencies
+      run: |
+        pip install build twine
+    - name: Build package
+      run: python -m build
+    - name: Publish to PyPI
+      env:
+        TWINE_USERNAME: __token__
+        TWINE_PASSWORD: ${{ secrets.PYPI_TOKEN }}
+      run: twine upload dist/*
+```
+
+## Best Practices
+
+1. **Version Bumping**: Use semantic versioning (MAJOR.MINOR.PATCH)
+2. **Testing**: Always test on Test PyPI before publishing to PyPI
+3. **Documentation**: Update README.md and CHANGELOG.md for each release
+4. **Git Tags**: Tag releases in git for easy reference
+5. **Dependencies**: Keep requirements.txt updated and specify version ranges
+
+## Resources
+
+- [Python Packaging Guide](https://packaging.python.org/)
+- [setuptools Documentation](https://setuptools.pypa.io/)
+- [PyPI Publishing Guide](https://packaging.python.org/tutorials/packaging-projects/)

INSTALL_TEST.md

@@ -0,0 +1,236 @@
+# Installation Testing Guide
+
+This guide helps you verify that the package installation works correctly.
+
+## Clean Installation Test
+
+### 1. Remove Any Previous Installations
+
+```bash
+# Deactivate any active virtual environment
+deactivate
+
+# Remove old virtual environment (if exists)
+rm -rf venv
+
+# Create fresh virtual environment
+python3 -m venv venv
+source venv/bin/activate  # On Linux/Mac
+# or
+venv\Scripts\activate  # On Windows
+```
+
+### 2. Install the Package
+
+#### Option A: Editable/Development Install
+
+```bash
+pip install -e .
+```
+
+This allows you to modify source code and see changes immediately.
+
+#### Option B: Regular Install
+
+```bash
+pip install .
+```
+
+This installs the package as if it were from PyPI.
+
+### 3. Verify Installation
+
+```bash
+# Check package is installed
+pip list | grep microscopy
+
+# Check version
+microscopy-detect --version
+# Expected output: microscopy-object-detection 1.0.0
+
+# Test Python import
+python -c "import src; print(src.__version__)"
+# Expected output: 1.0.0
+```
+
+### 4. Test Entry Points
+
+```bash
+# Test CLI
+microscopy-detect --help
+
+# Test GUI launcher (will open window)
+microscopy-detect-gui
+```
+
+### 5. Verify Package Contents
+
+```python
+# Run this in Python shell
+import src
+import src.database
+import src.model
+import src.gui
+
+# Check schema file is included
+from pathlib import Path
+import src.database
+db_path = Path(src.database.__file__).parent
+schema_file = db_path / 'schema.sql'
+print(f"Schema file exists: {schema_file.exists()}")
+# Expected: Schema file exists: True
+```
+
+## Troubleshooting
+
+### Issue: ModuleNotFoundError
+
+**Error:**
+```
+ModuleNotFoundError: No module named 'src'
+```
+
+**Solution:**
+```bash
+# Reinstall with verbose output
+pip install -e . -v
+
+# Or try regular install
+pip install . --force-reinstall
+```
+
+### Issue: Entry Points Not Working
+
+**Error:**
+```
+microscopy-detect: command not found
+```
+
+**Solution:**
+```bash
+# Check if scripts are in PATH
+which microscopy-detect
+
+# If not found, check pip install location
+pip show microscopy-object-detection
+
+# You might need to add to PATH or use full path
+~/.local/bin/microscopy-detect  # Linux
+```
+
+### Issue: Import Errors for PySide6
+
+**Error:**
+```
+ImportError: cannot import name 'QApplication' from 'PySide6.QtWidgets'
+```
+
+**Solution:**
+```bash
+# Install Qt dependencies (Linux only)
+sudo apt-get install libxcb-xinerama0
+
+# Reinstall PySide6
+pip uninstall PySide6
+pip install PySide6
+```
+
+### Issue: Config Files Not Found
+
+**Error:**
+```
+FileNotFoundError: config/app_config.yaml
+```
+
+**Solution:**
+The config file should be created automatically. If not:
+```bash
+# Create config directory in your home
+mkdir -p ~/.microscopy-detect
+cp config/app_config.yaml ~/.microscopy-detect/
+
+# Or run from source directory first time
+cd /home/martin/code/object_detection
+python main.py
+```
+
+## Manual Testing Checklist
+
+- [ ] Package installs without errors
+- [ ] Version command works (`microscopy-detect --version`)
+- [ ] Help command works (`microscopy-detect --help`)
+- [ ] GUI launches (`microscopy-detect-gui`)
+- [ ] Can import all modules in Python
+- [ ] Database schema file is accessible
+- [ ] Configuration loads correctly
+
+## Build and Install from Wheel
+
+```bash
+# Build the package
+python -m build
+
+# Install from wheel
+pip install dist/microscopy_object_detection-1.0.0-py3-none-any.whl
+
+# Test
+microscopy-detect --version
+```
+
+## Uninstall
+
+```bash
+pip uninstall microscopy-object-detection
+```
+
+## Development Workflow
+
+### After Code Changes
+
+If installed with `-e` (editable mode):
+- Python code changes are immediately available
+- No need to reinstall
+
+If installed with regular `pip install .`:
+- Reinstall after changes: `pip install . --force-reinstall`
+
+### After Adding New Files
+
+```bash
+# Reinstall to include new files
+pip install -e . --force-reinstall
+```
+
+## Expected Installation Output
+
+```
+Processing /home/martin/code/object_detection
+  Installing build dependencies ... done
+  Getting requirements to build wheel ... done
+  Preparing metadata (pyproject.toml) ... done
+Building wheels for collected packages: microscopy-object-detection
+  Building wheel for microscopy-object-detection (pyproject.toml) ... done
+Successfully built microscopy-object-detection
+Installing collected packages: microscopy-object-detection
+Successfully installed microscopy-object-detection-1.0.0
+```
+
+## Success Criteria
+
+Installation is successful when:
+1. ✅ No error messages during installation
+2. ✅ `pip list` shows the package
+3. ✅ `microscopy-detect --version` returns correct version
+4. ✅ GUI launches without errors
+5. ✅ All Python modules can be imported
+6. ✅ Database operations work
+7. ✅ Detection functionality works
+
+## Next Steps After Successful Install
+
+1. Configure image repository path
+2. Run first detection
+3. Train a custom model
+4. Export results
+
+For usage instructions, see [QUICKSTART.md](QUICKSTART.md)

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Your Name
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

MANIFEST.in

@@ -0,0 +1,37 @@
+# Include documentation files
+include README.md
+include LICENSE
+include ARCHITECTURE.md
+include IMPLEMENTATION_GUIDE.md
+include QUICKSTART.md
+include PLAN_SUMMARY.md
+
+# Include requirements
+include requirements.txt
+
+# Include configuration files
+recursive-include config *.yaml
+recursive-include config *.yml
+
+# Include database schema
+recursive-include src/database *.sql
+
+# Include tests
+recursive-include tests *.py
+
+# Exclude compiled Python files
+global-exclude *.pyc
+global-exclude *.pyo
+global-exclude __pycache__
+global-exclude *.so
+global-exclude .DS_Store
+
+# Exclude git and IDE files
+global-exclude .git*
+global-exclude .vscode
+global-exclude .idea
+
+# Exclude build artifacts
+prune build
+prune dist
+prune *.egg-info

PLAN_SUMMARY.md

@@ -0,0 +1,323 @@
+# Project Plan Summary - Microscopy Object Detection Application
+
+## Executive Summary
+
+This document provides a high-level overview of the planned microscopy object detection application. The application will enable users to train, validate, and deploy YOLOv8-based object detection models for microscopy images, with comprehensive GUI, database storage, and visualization capabilities.
+
+## Project Goals
+
+1. **Detect Objects**: Identify organelles and membrane branching structures in microscopy images
+2. **Train Models**: Fine-tune YOLOv8s on custom datasets with YOLO format annotations
+3. **Manage Results**: Store and query detection results in SQLite database
+4. **Visualize Data**: Interactive plots and image viewers for analysis
+5. **Export Results**: Flexible export options (CSV, JSON, Excel)
+6. **Future Annotation**: Provide manual annotation capabilities for dataset preparation
+
+## Key Technologies
+
+| Component | Technology | Purpose |
+|-----------|------------|---------|
+| ML Framework | Ultralytics YOLOv8 | Object detection model |
+| GUI | PySide6 | Desktop application interface |
+| Visualization | pyqtgraph | Real-time plots and charts |
+| Database | SQLite | Detection results storage |
+| Image Processing | OpenCV, Pillow | Image manipulation |
+
+## Application Architecture
+
+### Three-Layer Architecture
+
+```
+┌─────────────────────────────────────┐
+│         GUI Layer (PySide6)         │
+│  - Main Window with Tabs            │
+│  - Dialogs & Custom Widgets         │
+└──────────────┬──────────────────────┘
+               │
+┌──────────────▼──────────────────────┐
+│      Business Logic Layer           │
+│  - YOLO Wrapper                     │
+│  - Inference Engine                 │
+│  - Database Manager                 │
+│  - Config Manager                   │
+└──────────────┬──────────────────────┘
+               │
+┌──────────────▼──────────────────────┐
+│         Data Layer                  │
+│  - SQLite Database                  │
+│  - File System (images, models)     │
+│  - YOLOv8 Model Files               │
+└─────────────────────────────────────┘
+```
+
+## Core Features
+
+### 1. Detection Tab ✨
+- **Single Image Detection**: Select and detect objects in one image
+- **Batch Processing**: Process entire folders of images
+- **Real-time Preview**: View detections with bounding boxes
+- **Confidence Control**: Adjustable threshold slider
+- **Database Storage**: Automatic saving of results with metadata
+
+### 2. Training Tab 🎓
+- **Dataset Selection**: Browse for YOLO format datasets
+- **Parameter Configuration**: Epochs, batch size, image size, learning rate
+- **Progress Monitoring**: Real-time training metrics and loss curves
+- **Model Versioning**: Automatic model naming and version tracking
+- **Validation Metrics**: Track mAP, precision, recall during training
+
+### 3. Validation Tab 📊
+- **Model Evaluation**: Validate models on test datasets
+- **Metrics Visualization**: 
+  - Confusion matrix
+  - Precision-Recall curves
+  - Class-wise performance
+- **Model Comparison**: Compare multiple model versions
+
+### 4. Results Tab 📈
+- **Detection Browser**: Searchable table of all detections
+- **Advanced Filtering**: By date, model, class, confidence
+- **Statistics Dashboard**:
+  - Detection count by class
+  - Confidence distribution
+  - Timeline visualization
+- **Export Options**: CSV, JSON, Excel formats
+
+### 5. Annotation Tab 🖊️ (Future Feature)
+- **Image Browser**: Navigate through image repository
+- **Drawing Tools**: Create bounding boxes
+- **Class Assignment**: Label objects
+- **YOLO Export**: Export annotations for training
+
+## Database Design
+
+### Tables Overview
+
+| Table | Purpose | Key Fields |
+|-------|---------|------------|
+| **models** | Store trained models | name, version, path, metrics |
+| **images** | Image metadata | path, dimensions, checksum |
+| **detections** | Detection results | bbox, class, confidence |
+| **annotations** | Manual labels | bbox, class, annotator |
+
+### Key Relationships
+- Each detection links to one image and one model
+- Each image can have multiple detections from multiple models
+- Annotations are separate from automated detections
+
+## Implementation Phases
+
+### Phase 1: Core Foundation (Weeks 1-2)
+- [x] Architecture design and documentation
+- [ ] Project structure setup
+- [ ] Database schema implementation
+- [ ] YOLO wrapper basic functionality
+- [ ] Database manager CRUD operations
+
+### Phase 2: GUI Development (Weeks 3-4)
+- [ ] Main window and tab structure
+- [ ] Detection tab implementation
+- [ ] Training tab implementation
+- [ ] Basic visualization widgets
+- [ ] Configuration management
+
+### Phase 3: Advanced Features (Weeks 5-6)
+- [ ] Validation tab with metrics
+- [ ] Results tab with filtering
+- [ ] Export functionality
+- [ ] Error handling and logging
+- [ ] Performance optimization
+
+### Phase 4: Polish & Testing (Week 7)
+- [ ] Unit tests for all components
+- [ ] Integration testing
+- [ ] User documentation
+- [ ] Bug fixes and refinements
+- [ ] Deployment preparation
+
+### Phase 5: Future Enhancements (Post-Launch)
+- [ ] Annotation tool
+- [ ] Real-time camera detection
+- [ ] Multi-model ensemble
+- [ ] Cloud integration
+
+## File Structure
+
+```
+object_detection/
+├── main.py                    # Entry point
+├── requirements.txt           # Dependencies
+├── README.md                  # User documentation
+├── ARCHITECTURE.md            # Technical architecture
+├── IMPLEMENTATION_GUIDE.md    # Development guide
+├── PLAN_SUMMARY.md           # This file
+│
+├── config/
+│   └── app_config.yaml       # Application settings
+│
+├── src/
+│   ├── database/             # Database layer
+│   │   ├── db_manager.py     # Main database operations
+│   │   ├── models.py         # Data classes
+│   │   └── schema.sql        # SQL schema
+│   │
+│   ├── model/                # ML layer
+│   │   ├── yolo_wrapper.py   # YOLO interface
+│   │   └── inference.py      # Detection engine
+│   │
+│   ├── gui/                  # Presentation layer
+│   │   ├── main_window.py    # Main app window
+│   │   ├── tabs/             # Feature tabs
+│   │   ├── dialogs/          # Popup dialogs
+│   │   └── widgets/          # Custom widgets
+│   │
+│   └── utils/                # Utilities
+│       ├── config_manager.py # Config handling
+│       ├── logger.py         # Logging setup
+│       └── file_utils.py     # File operations
+│
+├── data/                     # Runtime data
+│   ├── models/               # Trained models
+│   ├── datasets/             # Training data
+│   └── results/              # Detection outputs
+│
+├── tests/                    # Test suite
+│   ├── test_database.py
+│   ├── test_model.py
+│   └── test_gui.py
+│
+├── logs/                     # Application logs
+└── docs/                     # Additional docs
+```
+
+## User Workflows
+
+### Workflow 1: First-Time Setup
+1. Launch application
+2. Configure image repository path in Settings
+3. Download/verify YOLOv8s.pt base model
+4. Ready to detect!
+
+### Workflow 2: Quick Detection
+1. Open Detection tab
+2. Select pre-trained model
+3. Choose image or folder
+4. Adjust confidence threshold
+5. Click "Detect"
+6. View results and save to database
+
+### Workflow 3: Train Custom Model
+1. Prepare dataset in YOLO format
+2. Open Training tab
+3. Select data.yaml file
+4. Configure hyperparameters
+5. Start training
+6. Monitor progress
+7. Model saved automatically with metrics
+
+### Workflow 4: Analyze Results
+1. Open Results tab
+2. Apply filters (date range, class, model)
+3. View statistics dashboard
+4. Export filtered results to CSV/JSON
+
+## Data Formats
+
+### YOLO Annotation Format
+```
+<class_id> <x_center> <y_center> <width> <height>
+```
+- All coordinates normalized (0-1)
+- One line per object
+- Separate .txt file for each image
+
+### Detection Output Format
+```json
+{
+  "image_id": 123,
+  "image_path": "relative/path/to/image.jpg",
+  "model_id": 5,
+  "model_name": "organelle_detector_v1",
+  "detections": [
+    {
+      "class_name": "organelle",
+      "confidence": 0.95,
+      "bbox": [0.1, 0.2, 0.3, 0.4],
+      "detected_at": "2024-12-04T15:30:00Z"
+    }
+  ]
+}
+```
+
+## Performance Requirements
+
+| Metric | Target | Notes |
+|--------|--------|-------|
+| Detection Speed | < 1s per image | On GPU for 640x640 images |
+| Training Start | < 5s | Model loading time |
+| Database Query | < 100ms | For 10k detections |
+| GUI Responsiveness | < 100ms | UI interactions |
+| Memory Usage | < 4GB | During inference |
+| Batch Processing | 100+ images | Without memory issues |
+
+## Risk Mitigation
+
+| Risk | Impact | Mitigation Strategy |
+|------|--------|-------------------|
+| CUDA/GPU issues | High | Fallback to CPU mode |
+| Database corruption | Medium | Regular backups, transactions |
+| Large datasets | Medium | Lazy loading, pagination |
+| Memory overflow | High | Batch size limits, monitoring |
+| Model compatibility | Low | Version checking, validation |
+
+## Success Criteria
+
+- ✅ Successfully detect objects in microscopy images
+- ✅ Train custom models with >80% mAP50
+- ✅ Store and retrieve detection results efficiently
+- ✅ Intuitive GUI requiring minimal training
+- ✅ Process 100+ images in batch mode
+- ✅ Export results in multiple formats
+- ✅ Comprehensive documentation
+
+## Next Steps
+
+1. **Review this plan** with stakeholders
+2. **Gather feedback** on features and priorities
+3. **Confirm requirements** are fully captured
+4. **Switch to Code mode** to begin implementation
+5. **Follow the implementation guide** for development
+6. **Iterate based on testing** and user feedback
+
+## Questions for Stakeholders
+
+Before proceeding to implementation, please consider:
+
+1. Are there any additional object classes beyond organelles and membrane branches?
+2. What is the typical size of your microscopy images?
+3. Do you need multi-user support or is single-user sufficient?
+4. Are there specific export formats you need beyond CSV/JSON/Excel?
+5. What is your expected dataset size (number of images)?
+6. Do you need support for 3D/volumetric microscopy images?
+7. Any specific visualization requirements?
+8. Integration with other tools or workflows?
+
+## Resources Required
+
+### Development Time
+- Phase 1-3: ~6 weeks development
+- Phase 4: ~1 week testing & polish
+- Total: ~7 weeks for initial release
+
+### Hardware Requirements
+- Development machine with CUDA GPU (recommended)
+- 16GB RAM minimum
+- 50GB storage for development
+
+### Software Dependencies
+- All listed in [`requirements.txt`](requirements.txt)
+- Total size: ~2GB when installed
+
+---
+
+**Ready to proceed with implementation? Let's build this! 🚀**

QUICKSTART.md

@@ -38,7 +38,7 @@ This will install:
 - OpenCV and Pillow (image processing)
 - And other dependencies
 
-**Note:** The first run will automatically download the YOLOv8s.pt model (~22MB).
+**Note:** The first run will automatically download the YOLOv8s-seg.pt segmentation model (~23MB).
 
 ### 4. Verify Installation
 
@@ -84,11 +84,11 @@ In the Settings dialog:
 ### Single Image Detection
 
 1. Go to the **Detection** tab
-2. Select a model from the dropdown (default: Base Model yolov8s.pt)
+2. Select a model from the dropdown (default: Base Model yolov8s-seg.pt)
 3. Adjust confidence threshold with the slider
 4. Click "Detect Single Image"
 5. Select an image file
-6. View results in the results panel
+6. View results with segmentation masks overlaid on the image
 
 ### Batch Detection
 
@@ -108,9 +108,18 @@ Detection results include:
 - **Class names**: Types of objects detected (e.g., organelle, membrane_branch)
 - **Confidence scores**: Detection confidence (0-1)
 - **Bounding boxes**: Object locations (stored in database)
+- **Segmentation masks**: Pixel-accurate polygon coordinates for each detected object
 
 All results are stored in the SQLite database at [`data/detections.db`](data/detections.db).
 
+### Segmentation Visualization
+
+The application automatically displays segmentation masks when available:
+- Semi-transparent colored overlay (30% opacity) showing the exact shape of detected objects
+- Polygon contours outlining each segmentation
+- Color-coded by object class
+- Toggle-able in future versions
+
 ## Database
 
 The application uses SQLite to store:
@@ -176,7 +185,7 @@ sudo apt-get install libxcb-xinerama0
 ### Detection Not Working
 
 **No models available**
-- The base YOLOv8s model will be downloaded automatically on first use
+- The base YOLOv8s-seg segmentation model will be downloaded automatically on first use
 - Make sure you have internet connection for the first run
 
 **Images not found**

README.md

@@ -1,6 +1,6 @@
 # Microscopy Object Detection Application
 
-A desktop application for detecting organelles and membrane branching structures in microscopy images using YOLOv8, featuring comprehensive training, validation, and visualization capabilities.
+A desktop application for detecting and segmenting organelles and membrane branching structures in microscopy images using YOLOv8-seg, featuring comprehensive training, validation, and visualization capabilities with pixel-accurate segmentation masks.
 
 ![Python](https://img.shields.io/badge/python-3.8+-blue.svg)
 ![PySide6](https://img.shields.io/badge/PySide6-6.5+-green.svg)
@@ -8,8 +8,8 @@ A desktop application for detecting organelles and membrane branching structures
 
 ## Features
 
-- **🎯 Object Detection**: Real-time and batch detection of microscopy objects
-- **🎓 Model Training**: Fine-tune YOLOv8s on custom microscopy datasets
+- **🎯 Object Detection & Segmentation**: Real-time and batch detection with pixel-accurate segmentation masks
+- **🎓 Model Training**: Fine-tune YOLOv8s-seg on custom microscopy datasets
 - **📊 Validation & Metrics**: Comprehensive model validation with visualization
 - **💾 Database Storage**: SQLite database for detection results and metadata
 - **📈 Visualization**: Interactive plots and charts using pyqtgraph
@@ -34,14 +34,24 @@ A desktop application for detecting organelles and membrane branching structures
 
 ## Installation
 
-### 1. Clone the Repository
+### Option 1: Install from PyPI (Recommended)
+
+```bash
+pip install microscopy-object-detection
+```
+
+This will install the package and all its dependencies.
+
+### Option 2: Install from Source
+
+#### 1. Clone the Repository
 
 ```bash
 git clone <repository-url>
 cd object_detection
 ```
 
-### 2. Create Virtual Environment
+#### 2. Create Virtual Environment
 
 ```bash
 # Linux/Mac
@@ -53,25 +63,44 @@ python -m venv venv
 venv\Scripts\activate
 ```
 
-### 3. Install Dependencies
+#### 3. Install in Development Mode
 
 ```bash
-pip install -r requirements.txt
+# Install in editable mode with dev dependencies
+pip install -e ".[dev]"
+
+# Or install just the package
+pip install .
 ```
 
 ### 4. Download Base Model
 
-The application will automatically download the YOLOv8s.pt model on first use, or you can download it manually:
+The application will automatically download the YOLOv8s-seg.pt segmentation model on first use, or you can download it manually:
 
 ```bash
 # The model will be downloaded automatically by ultralytics
 # Or download manually from: https://github.com/ultralytics/assets/releases
 ```
 
+**Note:** YOLOv8s-seg is a segmentation model that provides pixel-accurate masks for detected objects, enabling more precise analysis than standard bounding box detection.
+
 ## Quick Start
 
 ### 1. Launch the Application
 
+After installation, you can launch the application in two ways:
+
+**Using the GUI launcher:**
+```bash
+microscopy-detect-gui
+```
+
+**Or using Python directly:**
+```bash
+python -m microscopy_object_detection
+```
+
+**If installed from source:**
 ```bash
 python main.py
 ```
@@ -85,11 +114,12 @@ python main.py
 ### 3. Perform Detection
 
 1. Navigate to the **Detection** tab
-2. Select a model (default: yolov8s.pt)
+2. Select a model (default: yolov8s-seg.pt)
 3. Choose an image or folder
 4. Set confidence threshold
 5. Click **Detect**
-6. View results and save to database
+6. View results with segmentation masks overlaid
+7. Save results to database
 
 ### 4. Train Custom Model
 
@@ -212,8 +242,8 @@ The application uses SQLite with the following main tables:
 
 - **models**: Stores trained model information and metrics
 - **images**: Stores image metadata and paths
-- **detections**: Stores detection results with bounding boxes
-- **annotations**: Stores manual annotations (future feature)
+- **detections**: Stores detection results with bounding boxes and segmentation masks (polygon coordinates)
+- **annotations**: Stores manual annotations with optional segmentation masks (future feature)
 
 See [`ARCHITECTURE.md`](ARCHITECTURE.md) for detailed schema information.
 
@@ -230,7 +260,7 @@ image_repository:
   allowed_extensions: [".jpg", ".jpeg", ".png", ".tif", ".tiff"]
 
 models:
-  default_base_model: "yolov8s.pt"
+  default_base_model: "yolov8s-seg.pt"
   models_directory: "data/models"
 
 training:
@@ -258,7 +288,7 @@ visualization:
 from src.model.yolo_wrapper import YOLOWrapper
 
 # Initialize wrapper
-yolo = YOLOWrapper("yolov8s.pt")
+yolo = YOLOWrapper("yolov8s-seg.pt")
 
 # Train model
 results = yolo.train(
@@ -393,10 +423,10 @@ make html
 
 **Issue**: Model not found error
 
-**Solution**: Ensure YOLOv8s.pt is downloaded. Run:
+**Solution**: Ensure YOLOv8s-seg.pt is downloaded. Run:
 ```python
 from ultralytics import YOLO
-model = YOLO('yolov8s.pt')  # Will auto-download
+model = YOLO('yolov8s-seg.pt')  # Will auto-download
 ```
 
 

config/app_config.yaml

@@ -1,48 +1,57 @@
 database:
-  path: "data/detections.db"
-
+  path: data/detections.db
 image_repository:
-  base_path: ""  # Set by user through GUI
+  base_path: ''
   allowed_extensions:
-    - ".jpg"
-    - ".jpeg"
-    - ".png"
-    - ".tif"
-    - ".tiff"
-    - ".bmp"
-
+  - .jpg
+  - .jpeg
+  - .png
+  - .tif
+  - .tiff
+  - .bmp
 models:
-  default_base_model: "yolov8s.pt"
-  models_directory: "data/models"
-
+  default_base_model: yolov8s-seg.pt
+  models_directory: data/models
+  base_model_choices:
+  - yolov8s-seg.pt
+  - yolo11s-seg.pt
 training:
   default_epochs: 100
   default_batch_size: 16
-  default_imgsz: 640
+  default_imgsz: 1024
   default_patience: 50
   default_lr0: 0.01
-
+  two_stage:
+    enabled: false
+    stage1:
+      epochs: 20
+      lr0: 0.0005
+      patience: 10
+      freeze: 10
+    stage2:
+      epochs: 150
+      lr0: 0.0003
+      patience: 30
+  last_dataset_yaml: /home/martin/code/object_detection/data/datasets/data.yaml
+  last_dataset_dir: /home/martin/code/object_detection/data/datasets
 detection:
   default_confidence: 0.25
   default_iou: 0.45
   max_batch_size: 100
-
 visualization:
   bbox_colors:
-    organelle: "#FF6B6B"
-    membrane_branch: "#4ECDC4"
-    default: "#00FF00"
+    organelle: '#FF6B6B'
+    membrane_branch: '#4ECDC4'
+    default: '#00FF00'
   bbox_thickness: 2
   font_size: 12
-
 export:
   formats:
   - csv
   - json
   - excel
-  default_format: "csv"
-
+  default_format: csv
 logging:
-  level: "INFO"
-  file: "logs/app.log"
-  format: "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
+  level: INFO
+  file: logs/app.log
+  format: '%(asctime)s - %(name)s - %(levelname)s - %(message)s'

docs/16BIT_TIFF_SUPPORT.md

@@ -0,0 +1,300 @@
+# 16-bit TIFF Support for YOLO Object Detection
+
+## Overview
+
+This document describes the implementation of 16-bit grayscale TIFF support for YOLO object detection. The system properly loads 16-bit TIFF images, normalizes them to float32 [0-1], and handles them appropriately for both **inference** and **training** **without uint8 conversion** to preserve the full dynamic range and avoid data loss.
+
+## Key Features
+
+✅ Reads 16-bit or float32 images using tifffile
+✅ Converts to float32 [0-1] (NO uint8 conversion)
+✅ Replicates grayscale → RGB (3 channels)
+✅ **Inference**: Passes numpy arrays directly to YOLO (no file I/O)
+✅ **Training**: On-the-fly float32 conversion (NO disk caching)
+✅ Uses Ultralytics YOLOv8/v11 models
+✅ Works with segmentation models
+✅ No data loss, no double normalization, no silent clipping
+
+## Changes Made
+
+### 1. Dependencies ([`requirements.txt`](../requirements.txt:14))
+- Added `tifffile>=2023.0.0` for reliable 16-bit TIFF loading
+
+### 2. Image Loading ([`src/utils/image.py`](../src/utils/image.py))
+
+#### Enhanced TIFF Loading
+- Modified [`Image._load()`](../src/utils/image.py:87) to use `tifffile` for `.tif` and `.tiff` files
+- Preserves original 16-bit data type during loading
+- Properly handles both grayscale and multi-channel TIFF files
+
+#### New Normalization Method
+Added [`Image.to_normalized_float32()`](../src/utils/image.py:280) method that:
+- Converts image data to `float32`
+- Properly scales values to [0, 1] range:
+  - **16-bit images**: divides by 65535 (full dynamic range)
+  - 8-bit images: divides by 255
+  - Float images: clips to [0, 1]
+- Handles various data types automatically
+
+### 3. YOLO Preprocessing ([`src/model/yolo_wrapper.py`](../src/model/yolo_wrapper.py))
+
+Enhanced [`YOLOWrapper._prepare_source()`](../src/model/yolo_wrapper.py:231) to:
+1. Detect 16-bit TIFF files automatically
+2. Load and normalize to float32 [0-1] using the new method
+3. Replicate grayscale to RGB (3 channels)
+4. **Return numpy array directly** (NO file saving, NO uint8 conversion)
+5. Pass float32 array directly to YOLO for inference
+
+## Processing Pipeline
+
+### For Inference (predict)
+
+For 16-bit TIFF files during inference:
+
+1. **Load**: File loaded using `tifffile` → preserves 16-bit uint16 data
+2. **Normalize**: Convert to float32 and scale to [0, 1]
+   ```python
+   float_data = uint16_data.astype(np.float32) / 65535.0
+   ```
+3. **RGB Conversion**: Replicate grayscale to 3 channels
+   ```python
+   rgb_float = np.stack([float_data] * 3, axis=-1)
+   ```
+4. **Pass to YOLO**: Return float32 array directly (no uint8, no file I/O)
+5. **Inference**: YOLO processes the float32 [0-1] RGB array
+
+### For Training (train)
+
+Training now uses a custom dataset loader with on-the-fly conversion (NO disk caching):
+
+1. **Custom Dataset**: Uses `Float32Dataset` class that extends Ultralytics' `YOLODataset`
+2. **Load On-The-Fly**: Each image is loaded and converted during training:
+   - Detect 16-bit TIFF files automatically
+   - Load with `tifffile` (preserves uint16)
+   - Convert to float32 [0-1] in memory
+   - Replicate to 3 channels (RGB)
+3. **No Disk Cache**: Conversion happens in memory, no files written
+4. **Train**: YOLO trains on float32 [0-1] RGB arrays directly
+
+See [`src/utils/train_ultralytics_float.py`](../src/utils/train_ultralytics_float.py) for implementation.
+
+### No Data Loss!
+
+Unlike approaches that convert to uint8 (256 levels), this implementation:
+- Preserves full 16-bit dynamic range (65536 levels)
+- Maintains precision with float32 representation
+- For inference: passes data directly without file conversions
+- For training: uses float32 TIFFs (not uint8 PNGs)
+
+## Usage
+
+### Basic Image Loading
+
+```python
+from src.utils.image import Image
+
+# Load a 16-bit TIFF file
+img = Image("path/to/16bit_image.tif")
+
+# Get normalized float32 data [0-1]
+normalized = img.to_normalized_float32()  # Shape: (H, W), dtype: float32
+
+# Original data is preserved
+original = img.data  # Still uint16
+```
+
+### YOLO Inference
+
+The preprocessing is automatic - just use YOLO as normal:
+
+```python
+from src.model.yolo_wrapper import YOLOWrapper
+
+# Initialize model
+yolo = YOLOWrapper("yolov8s-seg.pt")
+yolo.load_model()
+
+# Perform inference on 16-bit TIFF
+# The image will be automatically normalized and passed as float32 [0-1]
+detections = yolo.predict("path/to/16bit_image.tif", conf=0.25)
+```
+
+### With InferenceEngine
+
+```python
+from src.model.inference import InferenceEngine
+from src.database.db_manager import DatabaseManager
+
+# Setup
+db = DatabaseManager("database.db")
+engine = InferenceEngine("model.pt", db, model_id=1)
+
+# Detect objects in 16-bit TIFF
+result = engine.detect_single(
+    image_path="path/to/16bit_image.tif",
+    relative_path="images/16bit_image.tif",
+    conf=0.25
+)
+```
+
+## Testing
+
+Three test scripts are provided:
+
+### 1. Image Loading Test
+```bash
+./venv/bin/python tests/test_16bit_tiff_loading.py
+```
+
+Tests:
+- Loading 16-bit TIFF files with tifffile
+- Normalization to float32 [0-1]
+- Data type and value range verification
+
+### 2. Float32 Passthrough Test (Most Important!)
+```bash
+./venv/bin/python tests/test_yolo_16bit_float32.py
+```
+
+Tests:
+- YOLO preprocessing returns numpy array (not file path)
+- Data is float32 [0-1] (not uint8)
+- No quantization to 256 levels (proves no uint8 conversion)
+- Sample output:
+  ```
+  ✓ SUCCESS: Prepared source is a numpy array (float32 passthrough)
+    Shape: (200, 200, 3)
+    Dtype: float32
+    Min value: 0.000000
+    Max value: 1.000000
+    Unique values: 399
+  
+  ✓ SUCCESS: Data has 399 unique values (> 256)
+    This confirms NO uint8 quantization occurred!
+  ```
+
+### 3. Legacy Test (Shows Old Behavior)
+```bash
+./venv/bin/python tests/test_yolo_16bit_preprocessing.py
+```
+
+This test shows the old behavior (uint8 conversion) - kept for comparison.
+
+## Benefits
+
+1. **No Data Loss**: Preserves full 16-bit dynamic range (65536 levels vs 256)
+2. **High Precision**: Float32 maintains fine-grained intensity differences
+3. **Automatic Processing**: No manual preprocessing needed
+4. **YOLO Compatible**: Ultralytics YOLO accepts float32 [0-1] arrays
+5. **Performance**: No intermediate file I/O for 16-bit TIFFs
+6. **Backwards Compatible**: Regular images (8-bit PNG, JPEG, etc.) still work as before
+
+## Technical Notes
+
+### Float32 vs uint8
+
+**With uint8 conversion (OLD - BAD):**
+- 16-bit (65536 levels) → uint8 (256 levels) = **99.6% data loss!**
+- Fine intensity differences are lost
+- Quantization artifacts
+
+**With float32 [0-1] (NEW - GOOD):**
+- 16-bit (65536 levels) → float32 (continuous) = **No data loss**
+- Full dynamic range preserved
+- Smooth gradients maintained
+
+### Memory Considerations
+
+For a 2048×2048 single-channel image:
+
+| Format | Memory | Disk Space | Notes |
+|--------|--------|------------|-------|
+| Original 16-bit | 8 MB | ~8 MB | uint16 grayscale TIFF |
+| Float32 grayscale | 16 MB | - | Intermediate |
+| Float32 3-channel | 48 MB | ~48 MB | Training cache |
+| uint8 RGB (old) | 12 MB | ~12 MB | OLD approach with data loss |
+
+The float32 approach uses ~3× more memory than uint8 during training but preserves **all information**.
+
+**No Disk Cache**: The new on-the-fly approach eliminates the need for cached datasets on disk.
+
+### Why Direct Numpy Array?
+
+Passing numpy arrays directly to YOLO (instead of saving to file):
+
+1. **Faster**: No disk I/O overhead
+2. **No Quantization**: Avoids PNG/JPEG quantization
+3. **Memory Efficient**: Single copy in memory
+4. **Cleaner**: No temp file management
+
+Ultralytics YOLO supports various input types:
+- File paths (str): `"image.jpg"`
+- Numpy arrays: `np.ndarray` ← **we use this**
+- PIL Images: `PIL.Image`
+- Torch tensors: `torch.Tensor`
+
+## Training with Float32 Dataset Loader
+
+The system now includes a custom dataset loader for 16-bit TIFF training:
+
+```python
+from src.utils.train_ultralytics_float import train_with_float32_loader
+
+# Train with on-the-fly float32 conversion
+results = train_with_float32_loader(
+    model_path="yolov8s-seg.pt",
+    data_yaml="data/my_dataset/data.yaml",
+    epochs=100,
+    batch=16,
+    imgsz=640,
+)
+```
+
+The `Float32Dataset` class automatically:
+- Detects 16-bit TIFF files
+- Loads with `tifffile` (not PIL/cv2)
+- Converts to float32 [0-1] on-the-fly
+- Replicates to 3 channels
+- Integrates seamlessly with Ultralytics training pipeline
+
+This is used automatically by the training tab in the GUI.
+
+## Installation
+
+Install the updated dependencies:
+
+```bash
+./venv/bin/pip install -r requirements.txt
+```
+
+Or install tifffile directly:
+
+```bash
+./venv/bin/pip install tifffile>=2023.0.0
+```
+
+## Example Test Output
+
+```
+=== Testing Float32 Passthrough (NO uint8) ===
+Created test 16-bit TIFF: /tmp/tmpdt5hm0ab.tif
+  Shape: (200, 200)
+  Dtype: uint16
+  Min value: 0
+  Max value: 65535
+
+Preprocessing result:
+  Prepared source type: <class 'numpy.ndarray'>
+
+✓ SUCCESS: Prepared source is a numpy array (float32 passthrough)
+  Shape: (200, 200, 3)
+  Dtype: float32
+  Min value: 0.000000
+  Max value: 1.000000
+  Mean value: 0.499992
+  Unique values: 399
+
+✓ SUCCESS: Data has 399 unique values (> 256)
+  This confirms NO uint8 quantization occurred!
+
+✓ All float32 passthrough tests passed!

docs/IMAGE_CLASS_USAGE.md

@@ -0,0 +1,220 @@
+# Image Class Usage Guide
+
+The `Image` class provides a convenient way to load and work with images in the microscopy object detection application.
+
+## Supported Formats
+
+The Image class supports the following image formats:
+- `.jpg`, `.jpeg` - JPEG images
+- `.png` - PNG images
+- `.tif`, `.tiff` - TIFF images (commonly used in microscopy)
+- `.bmp` - Bitmap images
+
+## Basic Usage
+
+### Loading an Image
+
+```python
+from src.utils import Image, ImageLoadError
+
+# Load an image from a file path
+try:
+    img = Image("path/to/image.jpg")
+    print(f"Loaded image: {img.width}x{img.height} pixels")
+except ImageLoadError as e:
+    print(f"Failed to load image: {e}")
+```
+
+### Accessing Image Properties
+
+```python
+img = Image("microscopy_image.tif")
+
+# Basic properties
+print(f"Width: {img.width} pixels")
+print(f"Height: {img.height} pixels")
+print(f"Channels: {img.channels}")
+print(f"Format: {img.format}")
+print(f"Shape: {img.shape}")  # (height, width, channels)
+
+# File information
+print(f"File size: {img.size_mb:.2f} MB")
+print(f"File size: {img.size_bytes} bytes")
+
+# Image type checks
+print(f"Is color: {img.is_color()}")
+print(f"Is grayscale: {img.is_grayscale()}")
+
+# String representation
+print(img)  # Shows summary of image properties
+```
+
+### Working with Image Data
+
+```python
+import numpy as np
+
+img = Image("sample.png")
+
+# Get image data as numpy array (OpenCV format, BGR)
+bgr_data = img.data
+print(f"Data shape: {bgr_data.shape}")
+print(f"Data type: {bgr_data.dtype}")
+
+# Get image as RGB (for display or processing)
+rgb_data = img.get_rgb()
+
+# Get grayscale version
+gray_data = img.get_grayscale()
+
+# Create a copy (for modifications)
+img_copy = img.copy()
+img_copy[0, 0] = [255, 255, 255]  # Modify copy, original unchanged
+
+# Resize image (returns new array, doesn't modify original)
+resized = img.resize(640, 640)
+```
+
+### Using PIL Image
+
+```python
+img = Image("photo.jpg")
+
+# Access as PIL Image (RGB format)
+pil_img = img.pil_image
+
+# Use PIL methods
+pil_img.show()  # Display image
+pil_img.save("output.png")  # Save with PIL
+```
+
+## Integration with YOLO
+
+```python
+from src.utils import Image
+from ultralytics import YOLO
+
+# Load model and image
+model = YOLO("yolov8n.pt")
+img = Image("microscopy/cell_01.tif")
+
+# Run inference (YOLO accepts file paths or numpy arrays)
+results = model(img.data)
+
+# Or use the file path directly
+results = model(str(img.path))
+```
+
+## Error Handling
+
+```python
+from src.utils import Image, ImageLoadError
+
+def process_image(image_path):
+    try:
+        img = Image(image_path)
+        # Process the image...
+        return img
+    except ImageLoadError as e:
+        print(f"Cannot load image: {e}")
+        return None
+```
+
+## Advanced Usage
+
+### Batch Processing
+
+```python
+from pathlib import Path
+from src.utils import Image, ImageLoadError
+
+def process_image_directory(directory):
+    """Process all images in a directory."""
+    image_paths = Path(directory).glob("*.tif")
+    
+    for path in image_paths:
+        try:
+            img = Image(path)
+            print(f"Processing {img.path.name}: {img.width}x{img.height}")
+            # Process the image...
+        except ImageLoadError as e:
+            print(f"Skipping {path}: {e}")
+```
+
+### Using with OpenCV Operations
+
+```python
+import cv2
+from src.utils import Image
+
+img = Image("input.jpg")
+
+# Apply OpenCV operations on the data
+blurred = cv2.GaussianBlur(img.data, (5, 5), 0)
+edges = cv2.Canny(img.data, 100, 200)
+
+# Note: These operations don't modify the original img.data
+```
+
+### Memory Efficient Processing
+
+```python
+from src.utils import Image
+
+# The Image class loads data into memory
+img = Image("large_image.tif")
+print(f"Image size in memory: {img.data.nbytes / (1024**2):.2f} MB")
+
+# When processing many images, consider loading one at a time
+# and releasing memory by deleting the object
+del img
+```
+
+## Best Practices
+
+1. **Always use try-except** when loading images to handle errors gracefully
+2. **Check image properties** before processing to ensure compatibility
+3. **Use copy()** when you need to modify image data without affecting the original
+4. **Path objects work too** - The class accepts both strings and Path objects
+5. **Consider memory usage** when working with large images or batches
+
+## Example: Complete Workflow
+
+```python
+from src.utils import Image, ImageLoadError
+from src.utils.file_utils import get_image_files
+
+def analyze_microscopy_images(directory):
+    """Analyze all microscopy images in a directory."""
+    
+    # Get all image files
+    image_files = get_image_files(directory, recursive=True)
+    
+    results = []
+    for image_path in image_files:
+        try:
+            # Load image
+            img = Image(image_path)
+            
+            # Analyze
+            result = {
+                'filename': img.path.name,
+                'width': img.width,
+                'height': img.height,
+                'channels': img.channels,
+                'format': img.format,
+                'size_mb': img.size_mb,
+                'is_color': img.is_color()
+            }
+            
+            results.append(result)
+            print(f"✓ Analyzed {img.path.name}")
+            
+        except ImageLoadError as e:
+            print(f"✗ Failed to load {image_path}: {e}")
+    
+    return results
+
+# Run analysis
+results = analyze_microscopy_images("data/datasets/cells")
+print(f"\nProcessed {len(results)} images")

docs/TRAINING_16BIT_TIFF.md

@@ -0,0 +1,269 @@
+# Training YOLO with 16-bit TIFF Datasets
+
+## Quick Start
+
+If your dataset contains 16-bit grayscale TIFF files, the training tab will automatically:
+
+1. Detect 16-bit TIFF images in your dataset
+2. Convert them to float32 [0-1] RGB **on-the-fly** during training
+3. Train without any disk caching (memory-efficient)
+
+**No manual intervention or disk space needed!**
+
+## Why Float32 On-The-Fly Conversion?
+
+### The Problem
+
+YOLO's training expects:
+- 3-channel images (RGB)
+- Images loaded from disk by the dataloader
+
+16-bit grayscale TIFFs are:
+- 1-channel (grayscale)
+- Need to be converted to RGB format
+
+### The Solution
+
+**NEW APPROACH (Current)**: On-the-fly float32 conversion
+- Load 16-bit TIFF with `tifffile` (not PIL/cv2)
+- Convert uint16 [0-65535] → float32 [0-1] in memory
+- Replicate grayscale to 3 channels
+- Pass directly to YOLO training pipeline
+- **No disk caching required!**
+
+**OLD APPROACH (Deprecated)**: Disk caching
+- Created 16-bit RGB PNG cache files on disk
+- Required ~2x dataset size in disk space
+- Slower first training run
+
+## How It Works
+
+### Custom Dataset Loader
+
+The system uses a custom `Float32Dataset` class that extends Ultralytics' `YOLODataset`:
+
+```python
+from src.utils.train_ultralytics_float import Float32Dataset
+
+# This dataset loader:
+# 1. Intercepts image loading
+# 2. Detects 16-bit TIFFs
+# 3. Converts to float32 [0-1] RGB on-the-fly
+# 4. Passes to training pipeline
+```
+
+### Conversion Process
+
+For each 16-bit grayscale TIFF during training:
+
+```
+1. Load with tifffile → uint16 [0, 65535]
+2. Convert to float32 → img.astype(float32) / 65535.0
+3. Replicate to RGB → np.stack([img] * 3, axis=-1)
+4. Result: float32 [0, 1] RGB array, shape (H, W, 3)
+```
+
+### Memory vs Disk
+
+| Aspect | On-the-fly (NEW) | Disk Cache (OLD) |
+|--------|------------------|------------------|
+| Disk Space | Dataset size only | ~2× dataset size |
+| First Training | Fast | Slow (creates cache) |
+| Subsequent Training | Fast | Fast |
+| Data Loss | None | None |
+| Setup Required | None | Cache creation |
+
+## Data Preservation
+
+### Float32 Precision
+
+16-bit TIFF: 65,536 levels (0-65535)
+Float32: ~7 decimal digits precision
+
+**Conversion accuracy:**
+```python
+Original: 32768 (uint16, middle intensity)
+Float32: 32768 / 65535 = 0.50000763 (exact)
+```
+
+Full 16-bit precision is preserved in float32 representation.
+
+### Comparison to uint8
+
+| Approach | Precision Loss | Recommended |
+|----------|----------------|-------------|
+| **float32 [0-1]** | None | ✓ YES |
+| uint16 RGB | None | ✓ YES (but disk-heavy) |
+| uint8 | 99.6% data loss | ✗ NO |
+
+**Why NO uint8:**
+```
+Original values:     32768, 32769, 32770 (distinct)
+Converted to uint8:  128,   128,   128   (collapsed!)
+```
+
+Multiple 16-bit values collapse to the same uint8 value.
+
+## Training Tab Behavior
+
+When you click "Start Training" with a 16-bit TIFF dataset:
+
+```
+[01:23:45] Exported 150 annotations across 50 image(s).
+[01:23:45] Using Float32 on-the-fly loader for 16-bit TIFF support (no disk caching)
+[01:23:45] Starting training run 'my_model_v1' using yolov8s-seg.pt
+[01:23:46] Using Float32Dataset loader for 16-bit TIFF support
+```
+
+Every training run uses the same approach - fast and efficient!
+
+## Inference vs Training
+
+| Operation | Input | Processing | Output to YOLO |
+|-----------|-------|------------|----------------|
+| **Inference** | 16-bit TIFF file | Load → float32 [0-1] → 3ch | numpy array (float32) |
+| **Training** | 16-bit TIFF dataset | Load on-the-fly → float32 [0-1] → 3ch | numpy array (float32) |
+
+Both preserve full 16-bit precision using float32 representation.
+
+## Technical Details
+
+### Custom Dataset Class
+
+Located in `src/utils/train_ultralytics_float.py`:
+
+```python
+class Float32Dataset(YOLODataset):
+    """
+    Extends Ultralytics YOLODataset to handle 16-bit TIFFs.
+    
+    Key methods:
+    - load_image(): Intercepts image loading
+    - Detects .tif/.tiff with dtype == uint16
+    - Converts: uint16 → float32 [0-1] → RGB (3-channel)
+    """
+```
+
+### Integration with YOLO
+
+The `YOLOWrapper.train()` method automatically uses the custom loader:
+
+```python
+# In src/model/yolo_wrapper.py
+def train(self, data_yaml, use_float32_loader=True, **kwargs):
+    if use_float32_loader:
+        # Use custom Float32Dataset
+        return train_with_float32_loader(...)
+    else:
+        # Standard YOLO training
+        return self.model.train(...)
+```
+
+### No PIL or cv2 for 16-bit
+
+16-bit TIFF loading uses `tifffile` directly:
+- PIL: Can load 16-bit but converts during processing
+- cv2: Limited 16-bit TIFF support
+- tifffile: Native 16-bit support, numpy output
+
+## Advantages Over Disk Caching
+
+### 1. No Disk Space Required
+```
+Dataset: 1000 images × 12 MB = 12 GB
+Old cache: Additional 24 GB (16-bit RGB PNGs)
+New approach: 0 GB additional (on-the-fly)
+```
+
+### 2. Faster Setup
+```
+Old: First training requires cache creation (minutes)
+New: Start training immediately (seconds)
+```
+
+### 3. Always In Sync
+```
+Old: Cache could become stale if images change
+New: Always loads current version from disk
+```
+
+### 4. Simpler Workflow
+```
+Old: Manage cache directory, cleanup, etc.
+New: Just point to dataset and train
+```
+
+## Troubleshooting
+
+### Error: "expected input to have 3 channels, but got 1"
+
+This shouldn't happen with the new Float32Dataset, but if it does:
+
+1. Check that `use_float32_loader=True` in training call
+2. Verify `Float32Dataset` is being used (check logs)
+3. Ensure `tifffile` is installed: `pip install tifffile`
+
+### Memory Usage
+
+On-the-fly conversion uses memory during training:
+- Image loaded: ~24 MB (2048×2048 uint16)
+- Converted float32 RGB: ~48 MB (temporary)
+- Released after augmentation pipeline
+
+**Mitigation:**
+- Reduce batch size if OOM errors occur
+- Images are processed one at a time during loading
+- Only active batch kept in memory
+
+### Slow Training
+
+If training seems slow:
+- Check disk I/O (slow disk can bottleneck loading)
+- Verify images aren't being re-converted each epoch (should cache after first load)
+- Monitor CPU usage during loading
+
+## Migration from Old Approach
+
+If you have existing cached datasets:
+
+```bash
+# Old cache location (safe to delete)
+rm -rf data/datasets/_float32_cache/
+
+# The new approach doesn't use this directory
+```
+
+Your original dataset structure remains unchanged:
+```
+data/my_dataset/
+├── train/
+│   ├── images/  (original 16-bit TIFFs)
+│   └── labels/
+├── val/
+│   ├── images/
+│   └── labels/
+└── data.yaml
+```
+
+Just point to the same `data.yaml` and train!
+
+## Performance Comparison
+
+| Metric | Old (Disk Cache) | New (On-the-fly) |
+|--------|------------------|------------------|
+| First training setup | 5-10 min | 0 sec |
+| Disk space overhead | 100% | 0% |
+| Training speed | Fast | Fast |
+| Subsequent runs | Fast | Fast |
+| Data accuracy | 16-bit preserved | 16-bit preserved |
+
+## Summary
+
+✓ **On-the-fly conversion**: Load and convert during training  
+✓ **No disk caching**: Zero additional disk space  
+✓ **Full precision**: Float32 preserves 16-bit dynamic range  
+✓ **No PIL/cv2**: Direct tifffile loading  
+✓ **Automatic**: Works transparently with training tab  
+✓ **Fast**: Efficient memory-based conversion  
+
+The new approach is simpler, faster to set up, and requires no disk space overhead!

examples/image_demo.py

@@ -0,0 +1,151 @@
+"""
+Example script demonstrating the Image class functionality.
+"""
+
+import sys
+from pathlib import Path
+
+# Add parent directory to path for imports
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+from src.utils import Image, ImageLoadError
+
+
+def demonstrate_image_loading():
+    """Demonstrate basic image loading functionality."""
+    print("=" * 60)
+    print("Image Class Demonstration")
+    print("=" * 60)
+
+    # Example 1: Try to load an image (replace with your own path)
+    example_paths = [
+        "data/datasets/example.jpg",
+        "data/datasets/sample.png",
+        "tests/test_image.jpg",
+    ]
+
+    loaded_img = None
+    for image_path in example_paths:
+        if Path(image_path).exists():
+            try:
+                print(f"\n1. Loading image: {image_path}")
+                img = Image(image_path)
+                loaded_img = img
+                print(f"   ✓ Successfully loaded!")
+                print(f"   {img}")
+                break
+            except ImageLoadError as e:
+                print(f"   ✗ Failed: {e}")
+        else:
+            print(f"\n1. Image not found: {image_path}")
+
+    if loaded_img is None:
+        print("\nNo example images found. Creating a test image...")
+        create_test_image()
+        return
+
+    # Example 2: Access image properties
+    print(f"\n2. Image Properties:")
+    print(f"   Width: {loaded_img.width} pixels")
+    print(f"   Height: {loaded_img.height} pixels")
+    print(f"   Channels: {loaded_img.channels}")
+    print(f"   Format: {loaded_img.format.upper()}")
+    print(f"   Shape: {loaded_img.shape}")
+    print(f"   File size: {loaded_img.size_mb:.2f} MB")
+    print(f"   Is color: {loaded_img.is_color()}")
+    print(f"   Is grayscale: {loaded_img.is_grayscale()}")
+
+    # Example 3: Get different formats
+    print(f"\n3. Accessing Image Data:")
+    print(f"   BGR data shape: {loaded_img.data.shape}")
+    print(f"   RGB data shape: {loaded_img.get_rgb().shape}")
+    print(f"   Grayscale shape: {loaded_img.get_grayscale().shape}")
+    print(f"   PIL image mode: {loaded_img.pil_image.mode}")
+
+    # Example 4: Resizing
+    print(f"\n4. Resizing Image:")
+    resized = loaded_img.resize(320, 320)
+    print(f"   Original size: {loaded_img.width}x{loaded_img.height}")
+    print(f"   Resized to: {resized.shape[1]}x{resized.shape[0]}")
+
+    # Example 5: Working with copies
+    print(f"\n5. Creating Copies:")
+    copy = loaded_img.copy()
+    print(f"   Created copy with shape: {copy.shape}")
+    print(f"   Original data unchanged: {(loaded_img.data == copy).all()}")
+
+    print("\n" + "=" * 60)
+    print("Demonstration Complete!")
+    print("=" * 60)
+
+
+def create_test_image():
+    """Create a test image for demonstration purposes."""
+    import cv2
+    import numpy as np
+
+    print("\nCreating a test image...")
+
+    # Create a colorful test image
+    width, height = 400, 300
+    test_img = np.zeros((height, width, 3), dtype=np.uint8)
+
+    # Add some colors
+    test_img[:100, :] = [255, 0, 0]  # Blue section
+    test_img[100:200, :] = [0, 255, 0]  # Green section
+    test_img[200:, :] = [0, 0, 255]  # Red section
+
+    # Save the test image
+    test_path = Path("test_demo_image.png")
+    cv2.imwrite(str(test_path), test_img)
+    print(f"Test image created: {test_path}")
+
+    # Now load and demonstrate with it
+    try:
+        img = Image(test_path)
+        print(f"\nLoaded test image: {img}")
+        print(f"Dimensions: {img.width}x{img.height}")
+        print(f"Channels: {img.channels}")
+        print(f"Format: {img.format}")
+
+        # Clean up
+        test_path.unlink()
+        print(f"\nTest image cleaned up.")
+
+    except ImageLoadError as e:
+        print(f"Error loading test image: {e}")
+
+
+def demonstrate_error_handling():
+    """Demonstrate error handling."""
+    print("\n" + "=" * 60)
+    print("Error Handling Demonstration")
+    print("=" * 60)
+
+    # Try to load non-existent file
+    print("\n1. Loading non-existent file:")
+    try:
+        img = Image("nonexistent.jpg")
+    except ImageLoadError as e:
+        print(f"   ✓ Caught error: {e}")
+
+    # Try unsupported format
+    print("\n2. Loading unsupported format:")
+    try:
+        # Create a text file
+        test_file = Path("test.txt")
+        test_file.write_text("not an image")
+        img = Image(test_file)
+    except ImageLoadError as e:
+        print(f"   ✓ Caught error: {e}")
+        test_file.unlink()  # Clean up
+
+    print("\n" + "=" * 60)
+
+
+if __name__ == "__main__":
+    print("\n")
+    demonstrate_image_loading()
+    print("\n")
+    demonstrate_error_handling()
+    print("\n")

main.py

@@ -0,0 +1,55 @@
+"""
+Microscopy Object Detection Application
+Main entry point for the application.
+"""
+
+import sys
+from pathlib import Path
+
+# Add src directory to path for development mode
+sys.path.insert(0, str(Path(__file__).parent))
+
+from PySide6.QtWidgets import QApplication
+from PySide6.QtCore import Qt
+
+from src import __version__
+from src.gui.main_window import MainWindow
+from src.utils.logger import setup_logging
+from src.utils.config_manager import ConfigManager
+
+
+def main():
+    """Application entry point."""
+    # Setup logging
+    config_manager = ConfigManager()
+    log_config = config_manager.get_section("logging")
+    setup_logging(
+        log_file=log_config.get("file", "logs/app.log"),
+        level=log_config.get("level", "INFO"),
+        log_format=log_config.get("format"),
+    )
+
+    # Enable High DPI scaling
+    QApplication.setHighDpiScaleFactorRoundingPolicy(
+        Qt.HighDpiScaleFactorRoundingPolicy.PassThrough
+    )
+
+    # Create Qt application
+    app = QApplication(sys.argv)
+    app.setApplicationName("Microscopy Object Detection")
+    app.setOrganizationName("MicroscopyLab")
+    app.setApplicationVersion(__version__)
+
+    # Set application style
+    app.setStyle("Fusion")
+
+    # Create and show main window
+    window = MainWindow()
+    window.show()
+
+    # Run application
+    sys.exit(app.exec())
+
+
+if __name__ == "__main__":
+    main()

pyproject.toml

@@ -0,0 +1,102 @@
+[build-system]
+requires = ["setuptools>=45", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "microscopy-object-detection"
+version = "1.0.0"
+description = "Desktop application for detecting and segmenting organelles in microscopy images using YOLOv8-seg"
+readme = "README.md"
+requires-python = ">=3.8"
+license = { text = "MIT" }
+authors = [{ name = "Your Name", email = "your.email@example.com" }]
+keywords = [
+    "microscopy",
+    "yolov8",
+    "object-detection",
+    "segmentation",
+    "computer-vision",
+    "deep-learning",
+]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Intended Audience :: Science/Research",
+    "Topic :: Scientific/Engineering :: Image Recognition",
+    "Topic :: Scientific/Engineering :: Bio-Informatics",
+    "License :: OSI Approved :: MIT License",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.8",
+    "Programming Language :: Python :: 3.9",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Operating System :: OS Independent",
+]
+
+dependencies = [
+    "ultralytics>=8.0.0",
+    "PySide6>=6.5.0",
+    "pyqtgraph>=0.13.0",
+    "numpy>=1.24.0",
+    "opencv-python>=4.8.0",
+    "Pillow>=10.0.0",
+    "PyYAML>=6.0",
+    "pandas>=2.0.0",
+    "openpyxl>=3.1.0",
+]
+
+[project.optional-dependencies]
+dev = [
+    "pytest>=7.0.0",
+    "pytest-cov>=4.0.0",
+    "black>=23.0.0",
+    "pylint>=2.17.0",
+    "mypy>=1.0.0",
+]
+
+[project.urls]
+Homepage = "https://github.com/yourusername/object_detection"
+Documentation = "https://github.com/yourusername/object_detection/blob/main/README.md"
+Repository = "https://github.com/yourusername/object_detection"
+"Bug Tracker" = "https://github.com/yourusername/object_detection/issues"
+
+[project.scripts]
+microscopy-detect = "src.cli:main"
+
+[project.gui-scripts]
+microscopy-detect-gui = "src.gui_launcher:main"
+
+[tool.setuptools]
+packages = [
+    "src",
+    "src.database",
+    "src.model",
+    "src.gui",
+    "src.gui.tabs",
+    "src.gui.dialogs",
+    "src.gui.widgets",
+    "src.utils",
+]
+include-package-data = true
+
+[tool.setuptools.package-data]
+"src.database" = ["*.sql"]
+
+[tool.black]
+line-length = 88
+target-version = ['py38', 'py39', 'py310', 'py311']
+include = '\.pyi?$'
+
+[tool.pylint.messages_control]
+max-line-length = 88
+
+[tool.mypy]
+python_version = "3.8"
+warn_return_any = true
+warn_unused_configs = true
+disallow_untyped_defs = false
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+python_files = ["test_*.py"]
+python_functions = ["test_*"]
+addopts = "-v --cov=src --cov-report=term-missing"

requirements.txt

@@ -11,6 +11,7 @@ pyqtgraph>=0.13.0
 opencv-python>=4.8.0
 Pillow>=10.0.0
 numpy>=1.24.0
+tifffile>=2023.0.0
 
 # Database
 sqlalchemy>=2.0.0

scripts/README_FLOAT32_TRAINING.md

@@ -0,0 +1,179 @@
+# Standalone Float32 Training Script for 16-bit TIFFs
+
+## Overview
+
+This standalone script (`train_float32_standalone.py`) trains YOLO models on 16-bit grayscale TIFF datasets with **no data loss**.
+
+- Loads 16-bit TIFFs with `tifffile` (not PIL/cv2)
+- Converts to float32 [0-1] on-the-fly (preserves full 16-bit precision)
+- Replicates grayscale → 3-channel RGB in memory
+- **No disk caching required**
+- Uses custom PyTorch Dataset + training loop
+
+## Quick Start
+
+```bash
+# Activate virtual environment
+source venv/bin/activate
+
+# Train on your 16-bit TIFF dataset
+python scripts/train_float32_standalone.py \
+    --data data/my_dataset/data.yaml \
+    --weights yolov8s-seg.pt \
+    --epochs 100 \
+    --batch 16 \
+    --imgsz 640 \
+    --lr 0.0001 \
+    --save-dir runs/my_training \
+    --device cuda
+```
+
+## Arguments
+
+| Argument | Required | Default | Description |
+|----------|----------|---------|-------------|
+| `--data` | Yes | - | Path to YOLO data.yaml file |
+| `--weights` | No | yolov8s-seg.pt | Pretrained model weights |
+| `--epochs` | No | 100 | Number of training epochs |
+| `--batch` | No | 16 | Batch size |
+| `--imgsz` | No | 640 | Input image size |
+| `--lr` | No | 0.0001 | Learning rate |
+| `--save-dir` | No | runs/train | Directory to save checkpoints |
+| `--device` | No | cuda/cpu | Training device (auto-detected) |
+
+## Dataset Format
+
+Your data.yaml should follow standard YOLO format:
+
+```yaml
+path: /path/to/dataset
+train: train/images
+val: val/images
+test: test/images  # optional
+
+names:
+  0: class1
+  1: class2
+
+nc: 2
+```
+
+Directory structure:
+```
+dataset/
+├── train/
+│   ├── images/
+│   │   ├── img1.tif  (16-bit grayscale TIFF)
+│   │   └── img2.tif
+│   └── labels/
+│       ├── img1.txt  (YOLO format)
+│       └── img2.txt
+├── val/
+│   ├── images/
+│   └── labels/
+└── data.yaml
+```
+
+## Output
+
+The script saves:
+- `epoch{N}.pt`: Checkpoint after each epoch
+- `best.pt`: Best model weights (lowest loss)
+- Training logs to console
+
+## Features
+
+✅ **16-bit precision preserved**: Float32 [0-1] maintains full dynamic range  
+✅ **No disk caching**: Conversion happens in memory  
+✅ **No PIL/cv2**: Direct tifffile loading  
+✅ **Variable-length labels**: Handles segmentation polygons  
+✅ **Checkpoint saving**: Resume training if interrupted  
+✅ **Best model tracking**: Automatically saves best weights  
+
+## Example
+
+Train a segmentation model on microscopy data:
+
+```bash
+python scripts/train_float32_standalone.py \
+    --data data/microscopy/data.yaml \
+    --weights yolov11s-seg.pt \
+    --epochs 150 \
+    --batch 8 \
+    --imgsz 1024 \
+    --lr 0.0003 \
+    --save-dir data/models/microscopy_v1
+```
+
+## Troubleshooting
+
+### Out of Memory (OOM)
+Reduce batch size:
+```bash
+--batch 4
+```
+
+### Slow Loading
+Reduce num_workers (edit script line 208):
+```python
+num_workers=2  # instead of 4
+```
+
+### Different Image Sizes
+The script expects all images to have the same dimensions. For variable sizes:
+1. Implement letterbox/resize in dataset's `_read_image()`
+2. Or preprocess images to same size
+
+### Loss Computation Errors
+If you see "Cannot determine loss", the script may need adjustment for your Ultralytics version. Check:
+```python
+# In train() function, the preds format may vary
+# Current script assumes: preds is tuple with loss OR dict with 'loss' key
+```
+
+## vs GUI Training
+
+| Feature | Standalone Script | GUI Training Tab |
+|---------|------------------|------------------|
+| Float32 conversion | ✓ Yes | ✓ Yes (automatic) |
+| Disk caching | ✗ None | ✗ None |
+| Progress UI | ✗ Console only | ✓ Visual progress bar |
+| Dataset selection | Manual CLI args | ✓ GUI browsing |
+| Multi-stage training | Manual runs | ✓ Built-in |
+| Use case | Advanced users | General users |
+
+## Technical Details
+
+### Data Loading Pipeline
+
+```
+16-bit TIFF file
+    ↓ (tifffile.imread)
+uint16 [0-65535]
+    ↓ (/ 65535.0)
+float32 [0-1]
+    ↓ (replicate channels)
+float32 RGB (H,W,3) [0-1]
+    ↓ (permute to C,H,W)
+torch.Tensor (3,H,W) float32
+    ↓ (DataLoader stack)
+Batch (B,3,H,W) float32
+    ↓
+YOLO Model
+```
+
+### Precision Comparison
+
+| Method | Unique Values | Data Loss |
+|--------|---------------|-----------|
+| **float32 [0-1]** | ~65,536 | None ✓ |
+| uint16 RGB | 65,536 | None ✓ |
+| uint8 | 256 | 99.6% ✗ |
+
+Example: Pixel value 32,768 (middle intensity)
+- Float32: 32768 / 65535.0 = 0.50000763 (exact)
+- uint8: 32768 → 128 → many values collapse!
+
+## License
+
+Same as main project.

scripts/train_float32_standalone.py

@@ -0,0 +1,351 @@
+#!/usr/bin/env python3
+"""
+Standalone training script for YOLO with 16-bit TIFF float32 support.
+
+This script trains YOLO models on 16-bit grayscale TIFF datasets without data loss.
+Converts images to float32 [0-1] on-the-fly using tifffile (no PIL/cv2).
+
+Usage:
+    python scripts/train_float32_standalone.py \\
+        --data path/to/data.yaml \\
+        --weights yolov8s-seg.pt \\
+        --epochs 100 \\
+        --batch 16 \\
+        --imgsz 640
+
+Based on the custom dataset approach to avoid Ultralytics' channel conversion issues.
+"""
+
+import argparse
+import os
+import sys
+import time
+from pathlib import Path
+
+import cv2
+import numpy as np
+import torch
+import torch.nn as nn
+import tifffile
+import yaml
+from torch.utils.data import Dataset, DataLoader
+from ultralytics import YOLO
+
+# Add project root to path
+project_root = Path(__file__).parent.parent
+sys.path.insert(0, str(project_root))
+
+from src.utils.logger import get_logger
+
+logger = get_logger(__name__)
+
+
+# ===================== Dataset =====================
+
+
+class Float32YOLODataset(Dataset):
+    """PyTorch dataset for 16-bit TIFF images with float32 conversion."""
+
+    def __init__(self, images_dir, labels_dir, img_size=640):
+        self.images_dir = Path(images_dir)
+        self.labels_dir = Path(labels_dir)
+        self.img_size = img_size
+
+        # Find images
+        extensions = {".tif", ".tiff", ".png", ".jpg", ".jpeg", ".bmp"}
+        self.paths = sorted(
+            [
+                p
+                for p in self.images_dir.rglob("*")
+                if p.is_file() and p.suffix.lower() in extensions
+            ]
+        )
+
+        if not self.paths:
+            raise ValueError(f"No images found in {images_dir}")
+
+        logger.info(f"Dataset: {len(self.paths)} images from {images_dir}")
+
+    def __len__(self):
+        return len(self.paths)
+
+    def _read_image(self, path: Path) -> np.ndarray:
+        """Load image as float32 [0-1] RGB."""
+        # Load with tifffile
+        img = tifffile.imread(str(path))
+
+        # Convert to float32
+        img = img.astype(np.float32)
+
+        # Normalize 16-bit→[0,1]
+        if img.max() > 1.5:
+            img = img / 65535.0
+
+        img = np.clip(img, 0.0, 1.0)
+
+        # Grayscale→RGB
+        if img.ndim == 2:
+            img = np.repeat(img[..., None], 3, axis=2)
+        elif img.ndim == 3 and img.shape[2] == 1:
+            img = np.repeat(img, 3, axis=2)
+
+        # Resize to model input size
+        img = cv2.resize(img, (self.img_size, self.img_size))
+
+        return img  # float32 (img_size, img_size, 3) [0,1] BGR
+
+    def _parse_label(self, path: Path) -> list:
+        """Parse YOLO label with variable-length  rows."""
+        if not path.exists():
+            return []
+
+        labels = []
+        with open(path, "r") as f:
+            for line in f:
+                vals = line.strip().split()
+                if len(vals) >= 5:
+                    labels.append([float(v) for v in vals])
+
+        return labels
+
+    def __getitem__(self, idx):
+        img_path = self.paths[idx]
+        label_path = self.labels_dir / f"{img_path.stem}.txt"
+
+        # Load & convert to tensor (C,H,W)
+        img = self._read_image(img_path)
+        img_t = torch.from_numpy(img).permute(2, 0, 1).contiguous()
+
+        # Load labels
+        labels = self._parse_label(label_path)
+
+        return img_t, labels, str(img_path.name)
+
+
+# ===================== Collate =====================
+
+
+def collate_fn(batch):
+    """Stack images, keep labels as list."""
+    imgs = torch.stack([b[0] for b in batch], dim=0)
+    labels = [b[1] for b in batch]
+    names = [b[2] for b in batch]
+    return imgs, labels, names
+
+
+# ===================== Training =====================
+
+
+def get_pytorch_model(ul_model):
+    """Extract PyTorch model and loss from Ultralytics wrapper."""
+    pt_model = None
+    loss_fn = None
+
+    # Try common patterns
+    if hasattr(ul_model, "model"):
+        pt_model = ul_model.model
+
+    # Find loss
+    if pt_model and hasattr(pt_model, "loss"):
+        loss_fn = pt_model.loss
+    elif pt_model and hasattr(pt_model, "compute_loss"):
+        loss_fn = pt_model.compute_loss
+
+    if pt_model is None:
+        raise RuntimeError("Could not extract PyTorch model")
+
+    return pt_model, loss_fn
+
+
+def train(args):
+    """Main training function."""
+    device = args.device
+    logger.info(f"Device: {device}")
+
+    # Parse data.yaml
+    with open(args.data, "r") as f:
+        data_config = yaml.safe_load(f)
+
+    dataset_root = Path(data_config.get("path", Path(args.data).parent))
+    train_img = dataset_root / data_config.get("train", "train/images")
+    val_img = dataset_root / data_config.get("val", "val/images")
+    train_lbl = train_img.parent / "labels"
+    val_lbl = val_img.parent / "labels"
+
+    # Load model
+    logger.info(f"Loading {args.weights}")
+    ul_model = YOLO(args.weights)
+    pt_model, loss_fn = get_pytorch_model(ul_model)
+
+    # Configure model args
+    from types import SimpleNamespace
+
+    if not hasattr(pt_model, "args"):
+        pt_model.args = SimpleNamespace()
+    if isinstance(pt_model.args, dict):
+        pt_model.args = SimpleNamespace(**pt_model.args)
+
+    # Set segmentation loss args
+    pt_model.args.overlap_mask = getattr(pt_model.args, "overlap_mask", True)
+    pt_model.args.mask_ratio = getattr(pt_model.args, "mask_ratio", 4)
+    pt_model.args.task = "segment"
+
+    pt_model.to(device)
+    pt_model.train()
+    for param in pt_model.parameters():
+        param.requires_grad = True
+
+    # Create datasets
+    train_ds = Float32YOLODataset(str(train_img), str(train_lbl), args.imgsz)
+    val_ds = Float32YOLODataset(str(val_img), str(val_lbl), args.imgsz)
+
+    train_loader = DataLoader(
+        train_ds,
+        batch_size=args.batch,
+        shuffle=True,
+        num_workers=4,
+        pin_memory=(device == "cuda"),
+        collate_fn=collate_fn,
+    )
+    val_loader = DataLoader(
+        val_ds,
+        batch_size=args.batch,
+        shuffle=False,
+        num_workers=2,
+        pin_memory=(device == "cuda"),
+        collate_fn=collate_fn,
+    )
+
+    # Optimizer
+    optimizer = torch.optim.AdamW(pt_model.parameters(), lr=args.lr)
+
+    # Training loop
+    os.makedirs(args.save_dir, exist_ok=True)
+    best_loss = float("inf")
+
+    for epoch in range(args.epochs):
+        t0 = time.time()
+        running_loss = 0.0
+        num_batches = 0
+
+        for imgs, labels_list, names in train_loader:
+            imgs = imgs.to(device)
+            optimizer.zero_grad()
+            num_batches += 1
+
+            # Forward (simple approach - just use preds)
+            preds = pt_model(imgs)
+
+            # Try to compute loss
+            # Simplest fallback: if preds is tuple/list, assume last element is loss
+            if isinstance(preds, (tuple, list)):
+                # Often YOLO forward returns (preds, loss) in training mode
+                if (
+                    len(preds) >= 2
+                    and isinstance(preds[-1], dict)
+                    and "loss" in preds[-1]
+                ):
+                    loss = preds[-1]["loss"]
+                elif len(preds) >= 2 and isinstance(preds[-1], torch.Tensor):
+                    loss = preds[-1]
+                else:
+                    # Manually compute using loss_fn if available
+                    if loss_fn:
+                        # This may fail - see logs
+                        try:
+                            loss_out = loss_fn(preds, labels_list)
+                            if isinstance(loss_out, dict):
+                                loss = loss_out["loss"]
+                            elif isinstance(loss_out, (tuple, list)):
+                                loss = loss_out[0]
+                            else:
+                                loss = loss_out
+                        except Exception as e:
+                            logger.error(f"Loss computation failed: {e}")
+                            logger.error(
+                                "Consider using Ultralytics .train() or check model/loss compatibility"
+                            )
+                            raise
+                    else:
+                        raise RuntimeError("Cannot determine loss from model output")
+            elif isinstance(preds, dict) and "loss" in preds:
+                loss = preds["loss"]
+            else:
+                raise RuntimeError(f"Unexpected preds format: {type(preds)}")
+
+            # Backward
+            loss = loss.mean()
+            loss.backward()
+            optimizer.step()
+
+            running_loss += loss.item()
+
+            if (num_batches % 10) == 0:
+                logger.info(
+                    f"Epoch {epoch+1} Batch {num_batches} Loss: {loss.item():.4f}"
+                )
+
+        epoch_loss = running_loss / max(1, num_batches)
+        epoch_time = time.time() - t0
+        logger.info(
+            f"Epoch {epoch+1}/{args.epochs} - Loss: {epoch_loss:.4f}, Time: {epoch_time:.1f}s"
+        )
+
+        # Save checkpoint
+        ckpt = Path(args.save_dir) / f"epoch{epoch+1}.pt"
+        torch.save(
+            {
+                "epoch": epoch + 1,
+                "model_state_dict": pt_model.state_dict(),
+                "optimizer_state_dict": optimizer.state_dict(),
+                "loss": epoch_loss,
+            },
+            ckpt,
+        )
+
+        # Save best
+        if epoch_loss < best_loss:
+            best_loss = epoch_loss
+            best_ckpt = Path(args.save_dir) / "best.pt"
+            torch.save(pt_model.state_dict(), best_ckpt)
+            logger.info(f"New best: {best_ckpt}")
+
+    logger.info("Training complete")
+
+
+# ===================== Main =====================
+
+
+def parse_args():
+    parser = argparse.ArgumentParser(
+        description="Train YOLO on 16-bit TIFF with float32"
+    )
+    parser.add_argument("--data", type=str, required=True, help="Path to data.yaml")
+    parser.add_argument(
+        "--weights", type=str, default="yolov8s-seg.pt", help="Pretrained weights"
+    )
+    parser.add_argument("--epochs", type=int, default=100, help="Number of epochs")
+    parser.add_argument("--batch", type=int, default=16, help="Batch size")
+    parser.add_argument("--imgsz", type=int, default=640, help="Image size")
+    parser.add_argument("--lr", type=float, default=1e-4, help="Learning rate")
+    parser.add_argument(
+        "--save-dir", type=str, default="runs/train", help="Save directory"
+    )
+    parser.add_argument(
+        "--device", type=str, default="cuda" if torch.cuda.is_available() else "cpu"
+    )
+    return parser.parse_args()
+
+
+if __name__ == "__main__":
+    args = parse_args()
+    logger.info("=" * 70)
+    logger.info("Float32 16-bit TIFF Training - Standalone Script")
+    logger.info("=" * 70)
+    logger.info(f"Data: {args.data}")
+    logger.info(f"Weights: {args.weights}")
+    logger.info(f"Epochs: {args.epochs}, Batch: {args.batch}, ImgSz: {args.imgsz}")
+    logger.info(f"LR: {args.lr}, Device: {args.device}")
+    logger.info("=" * 70)
+
+    train(args)

setup.py

@@ -0,0 +1,56 @@
+"""Setup script for Microscopy Object Detection Application."""
+
+from setuptools import setup, find_packages
+from pathlib import Path
+
+# Read the contents of README file
+this_directory = Path(__file__).parent
+long_description = (this_directory / "README.md").read_text(encoding="utf-8")
+
+# Read requirements
+requirements = (this_directory / "requirements.txt").read_text().splitlines()
+requirements = [
+    req.strip() for req in requirements if req.strip() and not req.startswith("#")
+]
+
+setup(
+    name="microscopy-object-detection",
+    version="1.0.0",
+    author="Your Name",
+    author_email="your.email@example.com",
+    description="Desktop application for detecting and segmenting organelles in microscopy images using YOLOv8-seg",
+    long_description=long_description,
+    long_description_content_type="text/markdown",
+    url="https://github.com/yourusername/object_detection",
+    packages=find_packages(exclude=["tests", "tests.*", "docs"]),
+    include_package_data=True,
+    install_requires=requirements,
+    python_requires=">=3.8",
+    classifiers=[
+        "Development Status :: 4 - Beta",
+        "Intended Audience :: Science/Research",
+        "Topic :: Scientific/Engineering :: Image Recognition",
+        "Topic :: Scientific/Engineering :: Bio-Informatics",
+        "License :: OSI Approved :: MIT License",
+        "Programming Language :: Python :: 3",
+        "Programming Language :: Python :: 3.8",
+        "Programming Language :: Python :: 3.9",
+        "Programming Language :: Python :: 3.10",
+        "Programming Language :: Python :: 3.11",
+        "Operating System :: OS Independent",
+    ],
+    entry_points={
+        "console_scripts": [
+            "microscopy-detect=src.cli:main",
+        ],
+        "gui_scripts": [
+            "microscopy-detect-gui=src.gui_launcher:main",
+        ],
+    },
+    keywords="microscopy yolov8 object-detection segmentation computer-vision deep-learning",
+    project_urls={
+        "Bug Reports": "https://github.com/yourusername/object_detection/issues",
+        "Source": "https://github.com/yourusername/object_detection",
+        "Documentation": "https://github.com/yourusername/object_detection/blob/main/README.md",
+    },
+)

src/__init__.py

@@ -0,0 +1,19 @@
+"""
+Microscopy Object Detection Application
+
+A desktop application for detecting and segmenting organelles and membrane
+branching structures in microscopy images using YOLOv8-seg.
+"""
+
+__version__ = "1.0.0"
+__author__ = "Your Name"
+__email__ = "your.email@example.com"
+__license__ = "MIT"
+
+# Package metadata
+__all__ = [
+    "__version__",
+    "__author__",
+    "__email__",
+    "__license__",
+]

src/cli.py

@@ -0,0 +1,61 @@
+"""
+Command-line interface for microscopy object detection application.
+"""
+
+import sys
+import argparse
+from pathlib import Path
+
+from src import __version__
+
+
+def main():
+    """Main CLI entry point."""
+    parser = argparse.ArgumentParser(
+        description="Microscopy Object Detection Application - CLI Interface",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  # Launch GUI
+  microscopy-detect-gui
+  
+  # Show version
+  microscopy-detect --version
+  
+  # Get help
+  microscopy-detect --help
+        """,
+    )
+
+    parser.add_argument(
+        "--version",
+        action="version",
+        version=f"microscopy-object-detection {__version__}",
+    )
+
+    parser.add_argument(
+        "--gui",
+        action="store_true",
+        help="Launch the GUI application (same as microscopy-detect-gui)",
+    )
+
+    args = parser.parse_args()
+
+    if args.gui:
+        # Launch GUI
+        try:
+            from src.gui_launcher import main as gui_main
+
+            gui_main()
+        except Exception as e:
+            print(f"Error launching GUI: {e}", file=sys.stderr)
+            sys.exit(1)
+    else:
+        # Show help if no arguments provided
+        parser.print_help()
+        print("\nTo launch the GUI, use: microscopy-detect-gui")
+        return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())

src/database/models.py

@@ -0,0 +1,69 @@
+"""
+Data models for the microscopy object detection application.
+These dataclasses represent the database entities.
+"""
+
+from dataclasses import dataclass
+from datetime import datetime
+from typing import Optional, Dict, Tuple, List
+
+
+@dataclass
+class Model:
+    """Represents a trained model."""
+
+    id: Optional[int]
+    model_name: str
+    model_version: str
+    model_path: str
+    base_model: str
+    created_at: datetime
+    training_params: Optional[Dict]
+    metrics: Optional[Dict]
+
+
+@dataclass
+class Image:
+    """Represents an image in the database."""
+
+    id: Optional[int]
+    relative_path: str
+    filename: str
+    width: int
+    height: int
+    captured_at: Optional[datetime]
+    added_at: datetime
+    checksum: Optional[str]
+
+
+@dataclass
+class Detection:
+    """Represents a detection result."""
+
+    id: Optional[int]
+    image_id: int
+    model_id: int
+    class_name: str
+    bbox: Tuple[float, float, float, float]  # (x_min, y_min, x_max, y_max)
+    confidence: float
+    segmentation_mask: Optional[
+        List[List[float]]
+    ]  # List of polygon coordinates [[x1,y1], [x2,y2], ...]
+    detected_at: datetime
+    metadata: Optional[Dict]
+
+
+@dataclass
+class Annotation:
+    """Represents a manual annotation."""
+
+    id: Optional[int]
+    image_id: int
+    class_name: str
+    bbox: Tuple[float, float, float, float]  # (x_min, y_min, x_max, y_max)
+    segmentation_mask: Optional[
+        List[List[float]]
+    ]  # List of polygon coordinates [[x1,y1], [x2,y2], ...]
+    annotator: str
+    created_at: datetime
+    verified: bool

src/database/schema.sql

@@ -0,0 +1,91 @@
+-- Microscopy Object Detection Application - Database Schema
+-- SQLite Database Schema for storing models, images, detections, and annotations
+
+-- Models table: stores trained model information
+CREATE TABLE IF NOT EXISTS models (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    model_name TEXT NOT NULL,
+    model_version TEXT NOT NULL,
+    model_path TEXT NOT NULL,
+    base_model TEXT NOT NULL DEFAULT 'yolov8s.pt',
+    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+    training_params TEXT,  -- JSON string of training parameters
+    metrics TEXT,          -- JSON string of validation metrics
+    UNIQUE(model_name, model_version)
+);
+
+-- Images table: stores image metadata
+CREATE TABLE IF NOT EXISTS images (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    relative_path TEXT NOT NULL UNIQUE,
+    filename TEXT NOT NULL,
+    width INTEGER,
+    height INTEGER,
+    captured_at TIMESTAMP,
+    added_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+    checksum TEXT
+);
+
+-- Detections table: stores detection results
+CREATE TABLE IF NOT EXISTS detections (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    image_id INTEGER NOT NULL,
+    model_id INTEGER NOT NULL,
+    class_name TEXT NOT NULL,
+    x_min REAL NOT NULL CHECK(x_min >= 0 AND x_min <= 1),
+    y_min REAL NOT NULL CHECK(y_min >= 0 AND y_min <= 1),
+    x_max REAL NOT NULL CHECK(x_max >= 0 AND x_max <= 1),
+    y_max REAL NOT NULL CHECK(y_max >= 0 AND y_max <= 1),
+    confidence REAL NOT NULL CHECK(confidence >= 0 AND confidence <= 1),
+    segmentation_mask TEXT,  -- JSON string of polygon coordinates [[x1,y1], [x2,y2], ...]
+    detected_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+    metadata TEXT,  -- JSON string for additional metadata
+    FOREIGN KEY (image_id) REFERENCES images (id) ON DELETE CASCADE,
+    FOREIGN KEY (model_id) REFERENCES models (id) ON DELETE CASCADE
+);
+
+-- Object classes table: stores annotation class definitions with colors
+CREATE TABLE IF NOT EXISTS object_classes (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    class_name TEXT NOT NULL UNIQUE,
+    color TEXT NOT NULL,  -- Hex color code (e.g., '#FF0000')
+    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+    description TEXT
+);
+
+-- Insert default object classes
+INSERT OR IGNORE INTO object_classes (class_name, color, description) VALUES
+    ('cell', '#FF0000', 'Cell object'),
+    ('nucleus', '#00FF00', 'Cell nucleus'),
+    ('mitochondria', '#0000FF', 'Mitochondria'),
+    ('vesicle', '#FFFF00', 'Vesicle');
+
+-- Annotations table: stores manual annotations
+CREATE TABLE IF NOT EXISTS annotations (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    image_id INTEGER NOT NULL,
+    class_id INTEGER NOT NULL,
+    x_min REAL NOT NULL CHECK(x_min >= 0 AND x_min <= 1),
+    y_min REAL NOT NULL CHECK(y_min >= 0 AND y_min <= 1),
+    x_max REAL NOT NULL CHECK(x_max >= 0 AND x_max <= 1),
+    y_max REAL NOT NULL CHECK(y_max >= 0 AND y_max <= 1),
+    segmentation_mask TEXT,  -- JSON string of polygon coordinates [[x1,y1], [x2,y2], ...]
+    annotator TEXT,
+    created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+    verified BOOLEAN DEFAULT 0,
+    FOREIGN KEY (image_id) REFERENCES images (id) ON DELETE CASCADE,
+    FOREIGN KEY (class_id) REFERENCES object_classes (id) ON DELETE CASCADE
+);
+
+-- Create indexes for performance optimization
+CREATE INDEX IF NOT EXISTS idx_detections_image_id ON detections(image_id);
+CREATE INDEX IF NOT EXISTS idx_detections_model_id ON detections(model_id);
+CREATE INDEX IF NOT EXISTS idx_detections_class_name ON detections(class_name);
+CREATE INDEX IF NOT EXISTS idx_detections_detected_at ON detections(detected_at);
+CREATE INDEX IF NOT EXISTS idx_detections_confidence ON detections(confidence);
+CREATE INDEX IF NOT EXISTS idx_images_relative_path ON images(relative_path);
+CREATE INDEX IF NOT EXISTS idx_images_added_at ON images(added_at);
+CREATE INDEX IF NOT EXISTS idx_annotations_image_id ON annotations(image_id);
+CREATE INDEX IF NOT EXISTS idx_annotations_class_id ON annotations(class_id);
+CREATE INDEX IF NOT EXISTS idx_models_created_at ON models(created_at);
+CREATE INDEX IF NOT EXISTS idx_object_classes_class_name ON object_classes(class_name);

src/gui/dialogs/config_dialog.py

@@ -0,0 +1,291 @@
+"""
+Configuration dialog for the microscopy object detection application.
+"""
+
+from PySide6.QtWidgets import (
+    QDialog,
+    QVBoxLayout,
+    QHBoxLayout,
+    QFormLayout,
+    QPushButton,
+    QLineEdit,
+    QSpinBox,
+    QDoubleSpinBox,
+    QFileDialog,
+    QTabWidget,
+    QWidget,
+    QLabel,
+    QGroupBox,
+)
+from PySide6.QtCore import Qt
+
+from src.utils.config_manager import ConfigManager
+from src.utils.logger import get_logger
+
+
+logger = get_logger(__name__)
+
+
+class ConfigDialog(QDialog):
+    """Configuration dialog window."""
+
+    def __init__(self, config_manager: ConfigManager, parent=None):
+        super().__init__(parent)
+
+        self.config_manager = config_manager
+
+        self.setWindowTitle("Settings")
+        self.setMinimumWidth(500)
+        self.setMinimumHeight(400)
+
+        self._setup_ui()
+        self._load_settings()
+
+    def _setup_ui(self):
+        """Setup user interface."""
+        layout = QVBoxLayout()
+
+        # Create tab widget for different setting categories
+        self.tab_widget = QTabWidget()
+
+        # General settings tab
+        general_tab = self._create_general_tab()
+        self.tab_widget.addTab(general_tab, "General")
+
+        # Training settings tab
+        training_tab = self._create_training_tab()
+        self.tab_widget.addTab(training_tab, "Training")
+
+        # Detection settings tab
+        detection_tab = self._create_detection_tab()
+        self.tab_widget.addTab(detection_tab, "Detection")
+
+        layout.addWidget(self.tab_widget)
+
+        # Buttons
+        button_layout = QHBoxLayout()
+        button_layout.addStretch()
+
+        self.save_button = QPushButton("Save")
+        self.save_button.clicked.connect(self.accept)
+        button_layout.addWidget(self.save_button)
+
+        self.cancel_button = QPushButton("Cancel")
+        self.cancel_button.clicked.connect(self.reject)
+        button_layout.addWidget(self.cancel_button)
+
+        layout.addLayout(button_layout)
+
+        self.setLayout(layout)
+
+    def _create_general_tab(self) -> QWidget:
+        """Create general settings tab."""
+        widget = QWidget()
+        layout = QVBoxLayout()
+
+        # Image repository group
+        repo_group = QGroupBox("Image Repository")
+        repo_layout = QFormLayout()
+
+        # Repository path
+        path_layout = QHBoxLayout()
+        self.repo_path_edit = QLineEdit()
+        self.repo_path_edit.setPlaceholderText("Path to image repository")
+        path_layout.addWidget(self.repo_path_edit)
+
+        browse_button = QPushButton("Browse...")
+        browse_button.clicked.connect(self._browse_repository)
+        path_layout.addWidget(browse_button)
+
+        repo_layout.addRow("Base Path:", path_layout)
+        repo_group.setLayout(repo_layout)
+        layout.addWidget(repo_group)
+
+        # Database group
+        db_group = QGroupBox("Database")
+        db_layout = QFormLayout()
+
+        self.db_path_edit = QLineEdit()
+        self.db_path_edit.setPlaceholderText("Path to database file")
+        db_layout.addRow("Database Path:", self.db_path_edit)
+
+        db_group.setLayout(db_layout)
+        layout.addWidget(db_group)
+
+        # Models group
+        models_group = QGroupBox("Models")
+        models_layout = QFormLayout()
+
+        self.models_dir_edit = QLineEdit()
+        self.models_dir_edit.setPlaceholderText("Directory for saved models")
+        models_layout.addRow("Models Directory:", self.models_dir_edit)
+
+        self.base_model_edit = QLineEdit()
+        self.base_model_edit.setPlaceholderText("yolov8s-seg.pt")
+        models_layout.addRow("Default Base Model:", self.base_model_edit)
+
+        models_group.setLayout(models_layout)
+        layout.addWidget(models_group)
+
+        layout.addStretch()
+        widget.setLayout(layout)
+        return widget
+
+    def _create_training_tab(self) -> QWidget:
+        """Create training settings tab."""
+        widget = QWidget()
+        layout = QVBoxLayout()
+
+        form_layout = QFormLayout()
+
+        # Epochs
+        self.epochs_spin = QSpinBox()
+        self.epochs_spin.setRange(1, 1000)
+        self.epochs_spin.setValue(100)
+        form_layout.addRow("Default Epochs:", self.epochs_spin)
+
+        # Batch size
+        self.batch_size_spin = QSpinBox()
+        self.batch_size_spin.setRange(1, 128)
+        self.batch_size_spin.setValue(16)
+        form_layout.addRow("Default Batch Size:", self.batch_size_spin)
+
+        # Image size
+        self.imgsz_spin = QSpinBox()
+        self.imgsz_spin.setRange(320, 1280)
+        self.imgsz_spin.setSingleStep(32)
+        self.imgsz_spin.setValue(640)
+        form_layout.addRow("Default Image Size:", self.imgsz_spin)
+
+        # Patience
+        self.patience_spin = QSpinBox()
+        self.patience_spin.setRange(1, 200)
+        self.patience_spin.setValue(50)
+        form_layout.addRow("Default Patience:", self.patience_spin)
+
+        # Learning rate
+        self.lr_spin = QDoubleSpinBox()
+        self.lr_spin.setRange(0.0001, 0.1)
+        self.lr_spin.setSingleStep(0.001)
+        self.lr_spin.setDecimals(4)
+        self.lr_spin.setValue(0.01)
+        form_layout.addRow("Default Learning Rate:", self.lr_spin)
+
+        layout.addLayout(form_layout)
+        layout.addStretch()
+        widget.setLayout(layout)
+        return widget
+
+    def _create_detection_tab(self) -> QWidget:
+        """Create detection settings tab."""
+        widget = QWidget()
+        layout = QVBoxLayout()
+
+        form_layout = QFormLayout()
+
+        # Confidence threshold
+        self.conf_spin = QDoubleSpinBox()
+        self.conf_spin.setRange(0.0, 1.0)
+        self.conf_spin.setSingleStep(0.05)
+        self.conf_spin.setDecimals(2)
+        self.conf_spin.setValue(0.25)
+        form_layout.addRow("Default Confidence:", self.conf_spin)
+
+        # IoU threshold
+        self.iou_spin = QDoubleSpinBox()
+        self.iou_spin.setRange(0.0, 1.0)
+        self.iou_spin.setSingleStep(0.05)
+        self.iou_spin.setDecimals(2)
+        self.iou_spin.setValue(0.45)
+        form_layout.addRow("Default IoU:", self.iou_spin)
+
+        # Max batch size
+        self.max_batch_spin = QSpinBox()
+        self.max_batch_spin.setRange(1, 1000)
+        self.max_batch_spin.setValue(100)
+        form_layout.addRow("Max Batch Size:", self.max_batch_spin)
+
+        layout.addLayout(form_layout)
+        layout.addStretch()
+        widget.setLayout(layout)
+        return widget
+
+    def _browse_repository(self):
+        """Browse for image repository directory."""
+        directory = QFileDialog.getExistingDirectory(
+            self, "Select Image Repository", self.repo_path_edit.text()
+        )
+
+        if directory:
+            self.repo_path_edit.setText(directory)
+
+    def _load_settings(self):
+        """Load current settings into dialog."""
+        # General settings
+        self.repo_path_edit.setText(
+            self.config_manager.get("image_repository.base_path", "")
+        )
+        self.db_path_edit.setText(
+            self.config_manager.get("database.path", "data/detections.db")
+        )
+        self.models_dir_edit.setText(
+            self.config_manager.get("models.models_directory", "data/models")
+        )
+        self.base_model_edit.setText(
+            self.config_manager.get("models.default_base_model", "yolov8s-seg.pt")
+        )
+
+        # Training settings
+        self.epochs_spin.setValue(
+            self.config_manager.get("training.default_epochs", 100)
+        )
+        self.batch_size_spin.setValue(
+            self.config_manager.get("training.default_batch_size", 16)
+        )
+        self.imgsz_spin.setValue(self.config_manager.get("training.default_imgsz", 640))
+        self.patience_spin.setValue(
+            self.config_manager.get("training.default_patience", 50)
+        )
+        self.lr_spin.setValue(self.config_manager.get("training.default_lr0", 0.01))
+
+        # Detection settings
+        self.conf_spin.setValue(
+            self.config_manager.get("detection.default_confidence", 0.25)
+        )
+        self.iou_spin.setValue(self.config_manager.get("detection.default_iou", 0.45))
+        self.max_batch_spin.setValue(
+            self.config_manager.get("detection.max_batch_size", 100)
+        )
+
+    def accept(self):
+        """Save settings and close dialog."""
+        logger.info("Saving configuration")
+
+        # Save general settings
+        self.config_manager.set(
+            "image_repository.base_path", self.repo_path_edit.text()
+        )
+        self.config_manager.set("database.path", self.db_path_edit.text())
+        self.config_manager.set("models.models_directory", self.models_dir_edit.text())
+        self.config_manager.set(
+            "models.default_base_model", self.base_model_edit.text()
+        )
+
+        # Save training settings
+        self.config_manager.set("training.default_epochs", self.epochs_spin.value())
+        self.config_manager.set(
+            "training.default_batch_size", self.batch_size_spin.value()
+        )
+        self.config_manager.set("training.default_imgsz", self.imgsz_spin.value())
+        self.config_manager.set("training.default_patience", self.patience_spin.value())
+        self.config_manager.set("training.default_lr0", self.lr_spin.value())
+
+        # Save detection settings
+        self.config_manager.set("detection.default_confidence", self.conf_spin.value())
+        self.config_manager.set("detection.default_iou", self.iou_spin.value())
+        self.config_manager.set("detection.max_batch_size", self.max_batch_spin.value())
+
+        # Save to file
+        self.config_manager.save_config()
+
+        super().accept()

src/gui/main_window.py

@@ -0,0 +1,309 @@
+"""
+Main window for the microscopy object detection application.
+"""
+
+from PySide6.QtWidgets import (
+    QMainWindow,
+    QTabWidget,
+    QMenuBar,
+    QMenu,
+    QStatusBar,
+    QMessageBox,
+    QWidget,
+    QVBoxLayout,
+    QLabel,
+)
+from PySide6.QtCore import Qt, QTimer, QSettings
+from PySide6.QtGui import QAction, QKeySequence
+
+from src.database.db_manager import DatabaseManager
+from src.utils.config_manager import ConfigManager
+from src.utils.logger import get_logger
+from src.gui.dialogs.config_dialog import ConfigDialog
+from src.gui.tabs.detection_tab import DetectionTab
+from src.gui.tabs.training_tab import TrainingTab
+from src.gui.tabs.validation_tab import ValidationTab
+from src.gui.tabs.results_tab import ResultsTab
+from src.gui.tabs.annotation_tab import AnnotationTab
+
+
+logger = get_logger(__name__)
+
+
+class MainWindow(QMainWindow):
+    """Main application window."""
+
+    def __init__(self):
+        super().__init__()
+
+        # Initialize managers
+        self.config_manager = ConfigManager()
+
+        db_path = self.config_manager.get_database_path()
+        self.db_manager = DatabaseManager(db_path)
+
+        logger.info("Main window initializing")
+
+        # Setup UI
+        self.setWindowTitle("Microscopy Object Detection")
+        self.setMinimumSize(1200, 800)
+
+        self._create_menu_bar()
+        self._create_tab_widget()
+        self._create_status_bar()
+
+        # Restore window geometry or center window on screen
+        self._restore_window_state()
+
+        logger.info("Main window initialized")
+
+    def _create_menu_bar(self):
+        """Create application menu bar."""
+        menubar = self.menuBar()
+
+        # File menu
+        file_menu = menubar.addMenu("&File")
+
+        settings_action = QAction("&Settings", self)
+        settings_action.setShortcut(QKeySequence("Ctrl+,"))
+        settings_action.triggered.connect(self._show_settings)
+        file_menu.addAction(settings_action)
+
+        file_menu.addSeparator()
+
+        exit_action = QAction("E&xit", self)
+        exit_action.setShortcut(QKeySequence("Ctrl+Q"))
+        exit_action.triggered.connect(self.close)
+        file_menu.addAction(exit_action)
+
+        # View menu
+        view_menu = menubar.addMenu("&View")
+
+        refresh_action = QAction("&Refresh", self)
+        refresh_action.setShortcut(QKeySequence("F5"))
+        refresh_action.triggered.connect(self._refresh_current_tab)
+        view_menu.addAction(refresh_action)
+
+        # Tools menu
+        tools_menu = menubar.addMenu("&Tools")
+
+        db_stats_action = QAction("Database &Statistics", self)
+        db_stats_action.triggered.connect(self._show_database_stats)
+        tools_menu.addAction(db_stats_action)
+
+        # Help menu
+        help_menu = menubar.addMenu("&Help")
+
+        about_action = QAction("&About", self)
+        about_action.triggered.connect(self._show_about)
+        help_menu.addAction(about_action)
+
+        docs_action = QAction("&Documentation", self)
+        docs_action.triggered.connect(self._show_documentation)
+        help_menu.addAction(docs_action)
+
+    def _create_tab_widget(self):
+        """Create main tab widget with all tabs."""
+        self.tab_widget = QTabWidget()
+        self.tab_widget.setTabPosition(QTabWidget.North)
+
+        # Create tabs
+        try:
+            self.detection_tab = DetectionTab(self.db_manager, self.config_manager)
+            self.training_tab = TrainingTab(self.db_manager, self.config_manager)
+            self.validation_tab = ValidationTab(self.db_manager, self.config_manager)
+            self.results_tab = ResultsTab(self.db_manager, self.config_manager)
+            self.annotation_tab = AnnotationTab(self.db_manager, self.config_manager)
+
+            # Add tabs to widget
+            self.tab_widget.addTab(self.detection_tab, "Detection")
+            self.tab_widget.addTab(self.training_tab, "Training")
+            self.tab_widget.addTab(self.validation_tab, "Validation")
+            self.tab_widget.addTab(self.results_tab, "Results")
+            self.tab_widget.addTab(self.annotation_tab, "Annotation (Future)")
+
+            # Connect tab change signal
+            self.tab_widget.currentChanged.connect(self._on_tab_changed)
+
+        except Exception as e:
+            logger.error(f"Error creating tabs: {e}")
+            # Create placeholder
+            placeholder = QWidget()
+            layout = QVBoxLayout()
+            layout.addWidget(QLabel(f"Error creating tabs: {e}"))
+            placeholder.setLayout(layout)
+            self.tab_widget.addTab(placeholder, "Error")
+
+        self.setCentralWidget(self.tab_widget)
+
+    def _create_status_bar(self):
+        """Create status bar."""
+        self.status_bar = QStatusBar()
+        self.setStatusBar(self.status_bar)
+
+        # Add permanent widgets to status bar
+        self.status_label = QLabel("Ready")
+        self.status_bar.addWidget(self.status_label)
+
+        # Initial status message
+        self._update_status("Ready")
+
+    def _center_window(self):
+        """Center window on screen."""
+        screen = self.screen().geometry()
+        size = self.geometry()
+        self.move(
+            (screen.width() - size.width()) // 2, (screen.height() - size.height()) // 2
+        )
+
+    def _restore_window_state(self):
+        """Restore window geometry from settings or center window."""
+        settings = QSettings("microscopy_app", "object_detection")
+        geometry = settings.value("main_window/geometry")
+
+        if geometry:
+            self.restoreGeometry(geometry)
+            logger.debug("Restored window geometry from settings")
+        else:
+            self._center_window()
+            logger.debug("Centered window on screen")
+
+    def _save_window_state(self):
+        """Save window geometry to settings."""
+        settings = QSettings("microscopy_app", "object_detection")
+        settings.setValue("main_window/geometry", self.saveGeometry())
+        logger.debug("Saved window geometry to settings")
+
+    def _show_settings(self):
+        """Show settings dialog."""
+        logger.info("Opening settings dialog")
+        dialog = ConfigDialog(self.config_manager, self)
+        if dialog.exec():
+            self._apply_settings()
+            self._update_status("Settings saved")
+
+    def _apply_settings(self):
+        """Apply changed settings."""
+        logger.info("Applying settings changes")
+        # Reload configuration in all tabs if needed
+        try:
+            if hasattr(self, "detection_tab"):
+                self.detection_tab.refresh()
+            if hasattr(self, "training_tab"):
+                self.training_tab.refresh()
+            if hasattr(self, "results_tab"):
+                self.results_tab.refresh()
+        except Exception as e:
+            logger.error(f"Error applying settings: {e}")
+
+    def _refresh_current_tab(self):
+        """Refresh the current tab."""
+        current_widget = self.tab_widget.currentWidget()
+        if hasattr(current_widget, "refresh"):
+            current_widget.refresh()
+            self._update_status("Tab refreshed")
+
+    def _on_tab_changed(self, index: int):
+        """Handle tab change event."""
+        tab_name = self.tab_widget.tabText(index)
+        logger.debug(f"Switched to tab: {tab_name}")
+        self._update_status(f"Viewing: {tab_name}")
+
+    def _show_database_stats(self):
+        """Show database statistics dialog."""
+        try:
+            stats = self.db_manager.get_detection_statistics()
+
+            message = f"""
+            <h3>Database Statistics</h3>
+            <p><b>Total Detections:</b> {stats.get('total_detections', 0)}</p>
+            <p><b>Average Confidence:</b> {stats.get('average_confidence', 0):.2%}</p>
+            <p><b>Classes:</b></p>
+            <ul>
+            """
+
+            for class_name, count in stats.get("class_counts", {}).items():
+                message += f"<li>{class_name}: {count}</li>"
+
+            message += "</ul>"
+
+            QMessageBox.information(self, "Database Statistics", message)
+
+        except Exception as e:
+            logger.error(f"Error getting database stats: {e}")
+            QMessageBox.warning(
+                self, "Error", f"Failed to get database statistics:\n{str(e)}"
+            )
+
+    def _show_about(self):
+        """Show about dialog."""
+        about_text = """
+        <h2>Microscopy Object Detection Application</h2>
+        <p><b>Version:</b> 1.0.0</p>
+        <p>A desktop application for detecting organelles and membrane branching 
+        structures in microscopy images using YOLOv8.</p>
+        
+        <p><b>Features:</b></p>
+        <ul>
+            <li>Object detection with YOLOv8</li>
+            <li>Model training and validation</li>
+            <li>Detection results storage</li>
+            <li>Interactive visualization</li>
+            <li>Export capabilities</li>
+        </ul>
+        
+        <p><b>Technologies:</b></p>
+        <ul>
+            <li>Ultralytics YOLOv8</li>
+            <li>PySide6</li>
+            <li>pyqtgraph</li>
+            <li>SQLite</li>
+        </ul>
+        """
+
+        QMessageBox.about(self, "About", about_text)
+
+    def _show_documentation(self):
+        """Show documentation."""
+        QMessageBox.information(
+            self,
+            "Documentation",
+            "Please refer to README.md and ARCHITECTURE.md files in the project directory.",
+        )
+
+    def _update_status(self, message: str, timeout: int = 5000):
+        """
+        Update status bar message.
+
+        Args:
+            message: Status message to display
+            timeout: Time in milliseconds to show message (0 for permanent)
+        """
+        self.status_label.setText(message)
+        if timeout > 0:
+            QTimer.singleShot(timeout, lambda: self.status_label.setText("Ready"))
+
+    def closeEvent(self, event):
+        """Handle window close event."""
+        reply = QMessageBox.question(
+            self,
+            "Confirm Exit",
+            "Are you sure you want to exit?",
+            QMessageBox.Yes | QMessageBox.No,
+            QMessageBox.No,
+        )
+
+        if reply == QMessageBox.Yes:
+            # Save window state before closing
+            self._save_window_state()
+
+            # Persist tab state and stop background work before exit
+            if hasattr(self, "training_tab"):
+                self.training_tab.shutdown()
+            if hasattr(self, "annotation_tab"):
+                self.annotation_tab.save_state()
+
+            logger.info("Application closing")
+            event.accept()
+        else:
+            event.ignore()

src/gui/tabs/annotation_tab.py

@@ -0,0 +1,566 @@
+"""
+Annotation tab for the microscopy object detection application.
+Manual annotation with pen tool and object class management.
+"""
+
+from PySide6.QtWidgets import (
+    QWidget,
+    QVBoxLayout,
+    QHBoxLayout,
+    QLabel,
+    QGroupBox,
+    QPushButton,
+    QFileDialog,
+    QMessageBox,
+    QSplitter,
+)
+from PySide6.QtCore import Qt, QSettings
+from pathlib import Path
+
+from src.database.db_manager import DatabaseManager
+from src.utils.config_manager import ConfigManager
+from src.utils.image import Image, ImageLoadError
+from src.utils.logger import get_logger
+from src.gui.widgets import AnnotationCanvasWidget, AnnotationToolsWidget
+
+logger = get_logger(__name__)
+
+
+class AnnotationTab(QWidget):
+    """Annotation tab for manual image annotation."""
+
+    def __init__(
+        self, db_manager: DatabaseManager, config_manager: ConfigManager, parent=None
+    ):
+        super().__init__(parent)
+        self.db_manager = db_manager
+        self.config_manager = config_manager
+        self.current_image = None
+        self.current_image_path = None
+        self.current_image_id = None
+        self.current_annotations = []
+        # IDs of annotations currently selected on the canvas (multi-select)
+        self.selected_annotation_ids = []
+
+        self._setup_ui()
+
+    def _setup_ui(self):
+        """Setup user interface."""
+        layout = QVBoxLayout()
+
+        # Main horizontal splitter to divide left (image) and right (controls)
+        self.main_splitter = QSplitter(Qt.Horizontal)
+        self.main_splitter.setHandleWidth(10)
+
+        # { Left splitter for image display and zoom info
+        self.left_splitter = QSplitter(Qt.Vertical)
+        self.left_splitter.setHandleWidth(10)
+
+        # Annotation canvas section
+        canvas_group = QGroupBox("Annotation Canvas")
+        canvas_layout = QVBoxLayout()
+
+        # Use the AnnotationCanvasWidget
+        self.annotation_canvas = AnnotationCanvasWidget()
+        self.annotation_canvas.zoom_changed.connect(self._on_zoom_changed)
+        self.annotation_canvas.annotation_drawn.connect(self._on_annotation_drawn)
+        # Selection of existing polylines (when tool is not in drawing mode)
+        self.annotation_canvas.annotation_selected.connect(self._on_annotation_selected)
+        canvas_layout.addWidget(self.annotation_canvas)
+
+        canvas_group.setLayout(canvas_layout)
+        self.left_splitter.addWidget(canvas_group)
+
+        # Controls info
+        controls_info = QLabel(
+            "Zoom: Mouse wheel or +/- keys | Drawing: Enable pen and drag mouse"
+        )
+        controls_info.setStyleSheet("QLabel { color: #888; font-style: italic; }")
+        self.left_splitter.addWidget(controls_info)
+        # }
+
+        # { Right splitter for annotation tools and controls
+        self.right_splitter = QSplitter(Qt.Vertical)
+        self.right_splitter.setHandleWidth(10)
+
+        # Annotation tools section
+        self.annotation_tools = AnnotationToolsWidget(self.db_manager)
+        self.annotation_tools.polyline_enabled_changed.connect(
+            self.annotation_canvas.set_polyline_enabled
+        )
+        self.annotation_tools.polyline_pen_color_changed.connect(
+            self.annotation_canvas.set_polyline_pen_color
+        )
+        self.annotation_tools.polyline_pen_width_changed.connect(
+            self.annotation_canvas.set_polyline_pen_width
+        )
+        # Show / hide bounding boxes
+        self.annotation_tools.show_bboxes_changed.connect(
+            self.annotation_canvas.set_show_bboxes
+        )
+        # RDP simplification controls
+        self.annotation_tools.simplify_on_finish_changed.connect(
+            self._on_simplify_on_finish_changed
+        )
+        self.annotation_tools.simplify_epsilon_changed.connect(
+            self._on_simplify_epsilon_changed
+        )
+        # Class selection and class-color changes
+        self.annotation_tools.class_selected.connect(self._on_class_selected)
+        self.annotation_tools.class_color_changed.connect(self._on_class_color_changed)
+        self.annotation_tools.clear_annotations_requested.connect(
+            self._on_clear_annotations
+        )
+        # Delete selected annotation on canvas
+        self.annotation_tools.delete_selected_annotation_requested.connect(
+            self._on_delete_selected_annotation
+        )
+        self.right_splitter.addWidget(self.annotation_tools)
+
+        # Image loading section
+        load_group = QGroupBox("Image Loading")
+        load_layout = QVBoxLayout()
+
+        # Load image button
+        button_layout = QHBoxLayout()
+        self.load_image_btn = QPushButton("Load Image")
+        self.load_image_btn.clicked.connect(self._load_image)
+        button_layout.addWidget(self.load_image_btn)
+        button_layout.addStretch()
+        load_layout.addLayout(button_layout)
+
+        # Image info label
+        self.image_info_label = QLabel("No image loaded")
+        load_layout.addWidget(self.image_info_label)
+
+        load_group.setLayout(load_layout)
+        self.right_splitter.addWidget(load_group)
+        # }
+
+        # Add both splitters to the main horizontal splitter
+        self.main_splitter.addWidget(self.left_splitter)
+        self.main_splitter.addWidget(self.right_splitter)
+
+        # Set initial sizes: 75% for left (image), 25% for right (controls)
+        self.main_splitter.setSizes([750, 250])
+
+        layout.addWidget(self.main_splitter)
+        self.setLayout(layout)
+
+        # Restore splitter positions from settings
+        self._restore_state()
+
+    def _load_image(self):
+        """Load and display an image file."""
+        # Get last opened directory from QSettings
+        settings = QSettings("microscopy_app", "object_detection")
+        last_dir = settings.value("annotation_tab/last_directory", None)
+
+        # Fallback to image repository path or home directory
+        if last_dir and Path(last_dir).exists():
+            start_dir = last_dir
+        else:
+            repo_path = self.config_manager.get_image_repository_path()
+            start_dir = repo_path if repo_path else str(Path.home())
+
+        # Open file dialog
+        file_path, _ = QFileDialog.getOpenFileName(
+            self,
+            "Select Image",
+            start_dir,
+            "Images (*" + " *".join(Image.SUPPORTED_EXTENSIONS) + ")",
+        )
+
+        if not file_path:
+            return
+
+        try:
+            # Load image using Image class
+            self.current_image = Image(file_path)
+            self.current_image_path = file_path
+
+            # Store the directory for next time
+            settings.setValue(
+                "annotation_tab/last_directory", str(Path(file_path).parent)
+            )
+
+            # Get or create image in database
+            relative_path = str(Path(file_path).name)  # Simplified for now
+            self.current_image_id = self.db_manager.get_or_create_image(
+                relative_path,
+                Path(file_path).name,
+                self.current_image.width,
+                self.current_image.height,
+            )
+
+            # Display image using the AnnotationCanvasWidget
+            self.annotation_canvas.load_image(self.current_image)
+
+            # Load and display any existing annotations for this image
+            self._load_annotations_for_current_image()
+
+            # Update info label
+            self._update_image_info()
+
+            logger.info(f"Loaded image: {file_path} (DB ID: {self.current_image_id})")
+
+        except ImageLoadError as e:
+            logger.error(f"Failed to load image: {e}")
+            QMessageBox.critical(
+                self, "Error Loading Image", f"Failed to load image:\n{str(e)}"
+            )
+        except Exception as e:
+            logger.error(f"Unexpected error loading image: {e}")
+            QMessageBox.critical(self, "Error", f"Unexpected error:\n{str(e)}")
+
+    def _update_image_info(self):
+        """Update the image info label with current image details."""
+        if self.current_image is None:
+            self.image_info_label.setText("No image loaded")
+            return
+
+        zoom_percentage = self.annotation_canvas.get_zoom_percentage()
+        info_text = (
+            f"File: {Path(self.current_image_path).name}\n"
+            f"Size: {self.current_image.width}x{self.current_image.height} pixels\n"
+            f"Channels: {self.current_image.channels}\n"
+            f"Data type: {self.current_image.dtype}\n"
+            f"Format: {self.current_image.format.upper()}\n"
+            f"File size: {self.current_image.size_mb:.2f} MB\n"
+            f"Zoom: {zoom_percentage}%"
+        )
+        self.image_info_label.setText(info_text)
+
+    def _on_zoom_changed(self, zoom_scale: float):
+        """Handle zoom level changes from the annotation canvas."""
+        self._update_image_info()
+
+    def _on_annotation_drawn(self, points: list):
+        """
+        Handle when an annotation stroke is drawn.
+
+        Saves the new annotation directly to the database and refreshes the
+        on-canvas display of annotations for the current image.
+        """
+        # Ensure we have an image loaded and in the DB
+        if not self.current_image or not self.current_image_id:
+            logger.warning("Annotation drawn but no image loaded")
+            QMessageBox.warning(
+                self,
+                "No Image",
+                "Please load an image before drawing annotations.",
+            )
+            return
+
+        current_class = self.annotation_tools.get_current_class()
+
+        if not current_class:
+            logger.warning("Annotation drawn but no object class selected")
+            QMessageBox.warning(
+                self,
+                "No Class Selected",
+                "Please select an object class before drawing annotations.",
+            )
+            return
+
+        if not points:
+            logger.warning("Annotation drawn with no points, ignoring")
+            return
+
+        # points are [(x_norm, y_norm), ...]
+        xs = [p[0] for p in points]
+        ys = [p[1] for p in points]
+        x_min, x_max = min(xs), max(xs)
+        y_min, y_max = min(ys), max(ys)
+
+        # Store segmentation mask in [y_norm, x_norm] format to match DB
+        db_polyline = [[float(y), float(x)] for (x, y) in points]
+
+        try:
+            annotation_id = self.db_manager.add_annotation(
+                image_id=self.current_image_id,
+                class_id=current_class["id"],
+                bbox=(x_min, y_min, x_max, y_max),
+                annotator="manual",
+                segmentation_mask=db_polyline,
+                verified=False,
+            )
+
+            logger.info(
+                f"Saved annotation (ID: {annotation_id}) for class "
+                f"'{current_class['class_name']}' "
+                f"Bounding box: ({x_min:.3f}, {y_min:.3f}) to ({x_max:.3f}, {y_max:.3f})\n"
+                f"with {len(points)} polyline points"
+            )
+
+            # Reload annotations from DB and redraw (respecting current class filter)
+            self._load_annotations_for_current_image()
+
+        except Exception as e:
+            logger.error(f"Failed to save annotation: {e}")
+            QMessageBox.critical(self, "Error", f"Failed to save annotation:\n{str(e)}")
+
+    def _on_annotation_selected(self, annotation_ids):
+        """
+        Handle selection of existing annotations on the canvas.
+
+        Args:
+            annotation_ids: List of selected annotation IDs, or None/empty if cleared.
+        """
+        if not annotation_ids:
+            self.selected_annotation_ids = []
+            self.annotation_tools.set_has_selected_annotation(False)
+            logger.debug("Annotation selection cleared on canvas")
+            return
+
+        # Normalize to a unique, sorted list of integer IDs
+        ids = sorted({int(aid) for aid in annotation_ids if isinstance(aid, int)})
+        self.selected_annotation_ids = ids
+        self.annotation_tools.set_has_selected_annotation(bool(ids))
+        logger.debug(f"Annotations selected on canvas: IDs={ids}")
+
+    def _on_simplify_on_finish_changed(self, enabled: bool):
+        """Update canvas simplify-on-finish flag from tools widget."""
+        self.annotation_canvas.simplify_on_finish = enabled
+        logger.debug(f"Annotation simplification on finish set to {enabled}")
+
+    def _on_simplify_epsilon_changed(self, epsilon: float):
+        """Update canvas RDP epsilon from tools widget."""
+        self.annotation_canvas.simplify_epsilon = float(epsilon)
+        logger.debug(f"Annotation simplification epsilon set to {epsilon}")
+
+    def _on_class_color_changed(self):
+        """
+        Handle changes to the selected object's class color.
+
+        When the user updates a class color in the tools widget, reload the
+        annotations for the current image so that all polylines are redrawn
+        using the updated per-class colors.
+        """
+        if not self.current_image_id:
+            return
+
+        logger.debug(
+            f"Class color changed; reloading annotations for image ID {self.current_image_id}"
+        )
+        self._load_annotations_for_current_image()
+
+    def _on_class_selected(self, class_data):
+        """
+        Handle when an object class is selected or cleared.
+
+        When a specific class is selected, only annotations of that class are drawn.
+        When the selection is cleared ("-- Select Class --"), all annotations are shown.
+        """
+        if class_data:
+            logger.debug(f"Object class selected: {class_data['class_name']}")
+        else:
+            logger.debug(
+                'No class selected ("-- Select Class --"), showing all annotations'
+            )
+
+        # Changing the class filter invalidates any previous selection
+        self.selected_annotation_ids = []
+        self.annotation_tools.set_has_selected_annotation(False)
+
+        # Whenever the selection changes, update which annotations are visible
+        self._redraw_annotations_for_current_filter()
+
+    def _on_clear_annotations(self):
+        """Handle clearing all annotations."""
+        self.annotation_canvas.clear_annotations()
+        # Clear in-memory state and selection, but keep DB entries unchanged
+        self.current_annotations = []
+        self.selected_annotation_ids = []
+        self.annotation_tools.set_has_selected_annotation(False)
+        logger.info("Cleared all annotations")
+
+    def _on_delete_selected_annotation(self):
+        """Handle deleting the currently selected annotation(s) (if any)."""
+        if not self.selected_annotation_ids:
+            QMessageBox.information(
+                self,
+                "No Selection",
+                "No annotation is currently selected.",
+            )
+            return
+
+        count = len(self.selected_annotation_ids)
+        if count == 1:
+            question = "Are you sure you want to delete the selected annotation?"
+            title = "Delete Annotation"
+        else:
+            question = (
+                f"Are you sure you want to delete the {count} selected annotations?"
+            )
+            title = "Delete Annotations"
+
+        reply = QMessageBox.question(
+            self,
+            title,
+            question,
+            QMessageBox.Yes | QMessageBox.No,
+            QMessageBox.No,
+        )
+        if reply != QMessageBox.Yes:
+            return
+
+        failed_ids = []
+        try:
+            for ann_id in self.selected_annotation_ids:
+                try:
+                    deleted = self.db_manager.delete_annotation(ann_id)
+                    if not deleted:
+                        failed_ids.append(ann_id)
+                except Exception as e:
+                    logger.error(f"Failed to delete annotation ID {ann_id}: {e}")
+                    failed_ids.append(ann_id)
+
+            if failed_ids:
+                QMessageBox.warning(
+                    self,
+                    "Partial Failure",
+                    "Some annotations could not be deleted:\n"
+                    + ", ".join(str(a) for a in failed_ids),
+                )
+            else:
+                logger.info(
+                    f"Deleted {count} annotation(s): "
+                    + ", ".join(str(a) for a in self.selected_annotation_ids)
+                )
+
+            # Clear selection and reload annotations for the current image from DB
+            self.selected_annotation_ids = []
+            self.annotation_tools.set_has_selected_annotation(False)
+            self._load_annotations_for_current_image()
+
+        except Exception as e:
+            logger.error(f"Failed to delete annotations: {e}")
+            QMessageBox.critical(
+                self,
+                "Error",
+                f"Failed to delete annotations:\n{str(e)}",
+            )
+
+    def _load_annotations_for_current_image(self):
+        """
+        Load all annotations for the current image from the database and
+        redraw them on the canvas, honoring the currently selected class
+        filter (if any).
+        """
+        if not self.current_image_id:
+            self.current_annotations = []
+            self.annotation_canvas.clear_annotations()
+            self.selected_annotation_ids = []
+            self.annotation_tools.set_has_selected_annotation(False)
+            return
+
+        try:
+            self.current_annotations = self.db_manager.get_annotations_for_image(
+                self.current_image_id
+            )
+            # New annotations loaded; reset any selection
+            self.selected_annotation_ids = []
+            self.annotation_tools.set_has_selected_annotation(False)
+            self._redraw_annotations_for_current_filter()
+        except Exception as e:
+            logger.error(
+                f"Failed to load annotations for image {self.current_image_id}: {e}"
+            )
+            QMessageBox.critical(
+                self,
+                "Error",
+                f"Failed to load annotations for this image:\n{str(e)}",
+            )
+
+    def _redraw_annotations_for_current_filter(self):
+        """
+        Redraw annotations for the current image, optionally filtered by the
+        currently selected object class.
+        """
+        # Clear current on-canvas annotations but keep the image
+        self.annotation_canvas.clear_annotations()
+
+        if not self.current_annotations:
+            return
+
+        current_class = self.annotation_tools.get_current_class()
+        selected_class_id = current_class["id"] if current_class else None
+
+        drawn_count = 0
+        for ann in self.current_annotations:
+            # Filter by class if one is selected
+            if (
+                selected_class_id is not None
+                and ann.get("class_id") != selected_class_id
+            ):
+                continue
+
+            if ann.get("segmentation_mask"):
+                polyline = ann["segmentation_mask"]
+                color = ann.get("class_color", "#FF0000")
+
+                self.annotation_canvas.draw_saved_polyline(
+                    polyline,
+                    color,
+                    width=3,
+                    annotation_id=ann["id"],
+                )
+                self.annotation_canvas.draw_saved_bbox(
+                    [ann["x_min"], ann["y_min"], ann["x_max"], ann["y_max"]],
+                    color,
+                    width=3,
+                )
+                drawn_count += 1
+
+        logger.info(
+            f"Displayed {drawn_count} annotation(s) for current image with "
+            f"{'no class filter' if selected_class_id is None else f'class_id={selected_class_id}'}"
+        )
+
+    def _restore_state(self):
+        """Restore splitter positions from settings."""
+        settings = QSettings("microscopy_app", "object_detection")
+
+        # Restore main splitter state
+        main_state = settings.value("annotation_tab/main_splitter_state")
+        if main_state:
+            self.main_splitter.restoreState(main_state)
+            logger.debug("Restored main splitter state")
+
+        # Restore left splitter state
+        left_state = settings.value("annotation_tab/left_splitter_state")
+        if left_state:
+            self.left_splitter.restoreState(left_state)
+            logger.debug("Restored left splitter state")
+
+        # Restore right splitter state
+        right_state = settings.value("annotation_tab/right_splitter_state")
+        if right_state:
+            self.right_splitter.restoreState(right_state)
+            logger.debug("Restored right splitter state")
+
+    def save_state(self):
+        """Save splitter positions to settings."""
+        settings = QSettings("microscopy_app", "object_detection")
+
+        # Save main splitter state
+        settings.setValue(
+            "annotation_tab/main_splitter_state", self.main_splitter.saveState()
+        )
+
+        # Save left splitter state
+        settings.setValue(
+            "annotation_tab/left_splitter_state", self.left_splitter.saveState()
+        )
+
+        # Save right splitter state
+        settings.setValue(
+            "annotation_tab/right_splitter_state", self.right_splitter.saveState()
+        )
+
+        logger.debug("Saved annotation tab splitter states")
+
+    def refresh(self):
+        """Refresh the tab."""
+        pass

src/gui/tabs/detection_tab.py

@@ -0,0 +1,466 @@
+"""
+Detection tab for the microscopy object detection application.
+Handles single image and batch detection.
+"""
+
+from PySide6.QtWidgets import (
+    QWidget,
+    QVBoxLayout,
+    QHBoxLayout,
+    QPushButton,
+    QLabel,
+    QComboBox,
+    QSlider,
+    QFileDialog,
+    QMessageBox,
+    QProgressBar,
+    QTextEdit,
+    QGroupBox,
+    QFormLayout,
+)
+from PySide6.QtCore import Qt, QThread, Signal
+from pathlib import Path
+from typing import Optional
+
+from src.database.db_manager import DatabaseManager
+from src.utils.config_manager import ConfigManager
+from src.utils.logger import get_logger
+from src.utils.file_utils import get_image_files
+from src.model.inference import InferenceEngine
+from src.utils.image import Image
+
+
+logger = get_logger(__name__)
+
+
+class DetectionWorker(QThread):
+    """Worker thread for running detection."""
+
+    progress = Signal(int, int, str)  # current, total, message
+    finished = Signal(list)  # results
+    error = Signal(str)  # error message
+
+    def __init__(self, engine, image_paths, repo_root, conf):
+        super().__init__()
+        self.engine = engine
+        self.image_paths = image_paths
+        self.repo_root = repo_root
+        self.conf = conf
+
+    def run(self):
+        """Run detection in background thread."""
+        try:
+            results = self.engine.detect_batch(
+                self.image_paths, self.repo_root, self.conf, self.progress.emit
+            )
+            self.finished.emit(results)
+        except Exception as e:
+            logger.error(f"Detection error: {e}")
+            self.error.emit(str(e))
+
+
+class DetectionTab(QWidget):
+    """Detection tab for single image and batch detection."""
+
+    def __init__(
+        self, db_manager: DatabaseManager, config_manager: ConfigManager, parent=None
+    ):
+        super().__init__(parent)
+        self.db_manager = db_manager
+        self.config_manager = config_manager
+        self.inference_engine = None
+        self.current_model_id = None
+
+        self._setup_ui()
+        self._connect_signals()
+        self._load_models()
+
+    def _setup_ui(self):
+        """Setup user interface."""
+        layout = QVBoxLayout()
+
+        # Model selection group
+        model_group = QGroupBox("Model Selection")
+        model_layout = QFormLayout()
+
+        self.model_combo = QComboBox()
+        self.model_combo.addItem("No models available", None)
+        model_layout.addRow("Model:", self.model_combo)
+
+        model_group.setLayout(model_layout)
+        layout.addWidget(model_group)
+
+        # Detection settings group
+        settings_group = QGroupBox("Detection Settings")
+        settings_layout = QFormLayout()
+
+        # Confidence threshold
+        conf_layout = QHBoxLayout()
+        self.conf_slider = QSlider(Qt.Horizontal)
+        self.conf_slider.setRange(0, 100)
+        self.conf_slider.setValue(25)
+        self.conf_slider.setTickPosition(QSlider.TicksBelow)
+        self.conf_slider.setTickInterval(10)
+        conf_layout.addWidget(self.conf_slider)
+
+        self.conf_label = QLabel("0.25")
+        conf_layout.addWidget(self.conf_label)
+
+        settings_layout.addRow("Confidence:", conf_layout)
+        settings_group.setLayout(settings_layout)
+        layout.addWidget(settings_group)
+
+        # Action buttons
+        button_layout = QHBoxLayout()
+
+        self.single_image_btn = QPushButton("Detect Single Image")
+        self.single_image_btn.clicked.connect(self._detect_single_image)
+        button_layout.addWidget(self.single_image_btn)
+
+        self.batch_btn = QPushButton("Detect Batch (Folder)")
+        self.batch_btn.clicked.connect(self._detect_batch)
+        button_layout.addWidget(self.batch_btn)
+
+        layout.addLayout(button_layout)
+
+        # Progress bar
+        self.progress_bar = QProgressBar()
+        self.progress_bar.setVisible(False)
+        layout.addWidget(self.progress_bar)
+
+        # Results display
+        results_group = QGroupBox("Detection Results")
+        results_layout = QVBoxLayout()
+
+        self.results_text = QTextEdit()
+        self.results_text.setReadOnly(True)
+        self.results_text.setMaximumHeight(200)
+        results_layout.addWidget(self.results_text)
+
+        results_group.setLayout(results_layout)
+        layout.addWidget(results_group)
+
+        layout.addStretch()
+        self.setLayout(layout)
+
+    def _connect_signals(self):
+        """Connect signals and slots."""
+        self.conf_slider.valueChanged.connect(self._update_confidence_label)
+        self.model_combo.currentIndexChanged.connect(self._on_model_changed)
+
+    def _load_models(self):
+        """Load available models from database and local storage."""
+        try:
+            self.model_combo.clear()
+            models = self.db_manager.get_models()
+            has_models = False
+
+            known_paths = set()
+
+            # Add base model option first (always available)
+            base_model = self.config_manager.get(
+                "models.default_base_model", "yolov8s-seg.pt"
+            )
+            if base_model:
+                base_data = {
+                    "id": 0,
+                    "path": base_model,
+                    "model_name": Path(base_model).stem or "Base Model",
+                    "model_version": "pretrained",
+                    "base_model": base_model,
+                    "source": "base",
+                }
+                self.model_combo.addItem(f"Base Model ({base_model})", base_data)
+                known_paths.add(self._normalize_model_path(base_model))
+                has_models = True
+
+            # Add trained models from database
+            for model in models:
+                display_name = f"{model['model_name']} v{model['model_version']}"
+                model_data = {**model, "path": model.get("model_path")}
+                normalized = self._normalize_model_path(model_data.get("path"))
+                if normalized:
+                    known_paths.add(normalized)
+                self.model_combo.addItem(display_name, model_data)
+                has_models = True
+
+            # Discover local model files not yet in the database
+            local_models = self._discover_local_models()
+            for model_path in local_models:
+                normalized = self._normalize_model_path(model_path)
+                if normalized in known_paths:
+                    continue
+
+                display_name = f"Local Model ({Path(model_path).stem})"
+                model_data = {
+                    "id": None,
+                    "path": str(model_path),
+                    "model_name": Path(model_path).stem,
+                    "model_version": "local",
+                    "base_model": Path(model_path).stem,
+                    "source": "local",
+                }
+                self.model_combo.addItem(display_name, model_data)
+                known_paths.add(normalized)
+                has_models = True
+
+            if not has_models:
+                self.model_combo.addItem("No models available", None)
+                self._set_buttons_enabled(False)
+            else:
+                self._set_buttons_enabled(True)
+
+        except Exception as e:
+            logger.error(f"Error loading models: {e}")
+            QMessageBox.warning(self, "Error", f"Failed to load models:\n{str(e)}")
+
+    def _on_model_changed(self, index: int):
+        """Handle model selection change."""
+        model_data = self.model_combo.itemData(index)
+        if model_data and model_data["id"] != 0:
+            self.current_model_id = model_data["id"]
+        else:
+            self.current_model_id = None
+
+    def _update_confidence_label(self, value: int):
+        """Update confidence label."""
+        conf = value / 100.0
+        self.conf_label.setText(f"{conf:.2f}")
+
+    def _detect_single_image(self):
+        """Detect objects in a single image."""
+        # Get image file
+        repo_path = self.config_manager.get_image_repository_path()
+        start_dir = repo_path if repo_path else ""
+
+        file_path, _ = QFileDialog.getOpenFileName(
+            self,
+            "Select Image",
+            start_dir,
+            "Images (*" + " *".join(Image.SUPPORTED_EXTENSIONS) + ")",
+        )
+
+        if not file_path:
+            return
+
+        # Run detection
+        self._run_detection([file_path])
+
+    def _detect_batch(self):
+        """Detect objects in batch (folder)."""
+        # Get folder
+        repo_path = self.config_manager.get_image_repository_path()
+        start_dir = repo_path if repo_path else ""
+
+        folder_path = QFileDialog.getExistingDirectory(self, "Select Folder", start_dir)
+
+        if not folder_path:
+            return
+
+        # Get all image files
+        allowed_ext = self.config_manager.get_allowed_extensions()
+        image_files = get_image_files(folder_path, allowed_ext, recursive=False)
+
+        if not image_files:
+            QMessageBox.information(
+                self, "No Images", "No image files found in selected folder."
+            )
+            return
+
+        # Confirm batch processing
+        reply = QMessageBox.question(
+            self,
+            "Confirm Batch Detection",
+            f"Process {len(image_files)} images?",
+            QMessageBox.Yes | QMessageBox.No,
+        )
+
+        if reply == QMessageBox.Yes:
+            self._run_detection(image_files)
+
+    def _run_detection(self, image_paths: list):
+        """Run detection on image list."""
+        try:
+            # Get selected model
+            model_data = self.model_combo.currentData()
+            if not model_data:
+                QMessageBox.warning(self, "No Model", "Please select a model first.")
+                return
+
+            model_path = model_data.get("path")
+            if not model_path:
+                QMessageBox.warning(
+                    self, "Invalid Model", "Selected model is missing a file path."
+                )
+                return
+
+            if not Path(model_path).exists():
+                QMessageBox.critical(
+                    self,
+                    "Model Not Found",
+                    f"The selected model file could not be found:\n{model_path}",
+                )
+                return
+
+            model_id = model_data.get("id")
+
+            # Ensure we have a database entry for the selected model
+            if model_id in (None, 0):
+                model_id = self._ensure_model_record(model_data)
+                if not model_id:
+                    QMessageBox.critical(
+                        self,
+                        "Model Registration Failed",
+                        "Unable to register the selected model in the database.",
+                    )
+                    return
+
+            normalized_model_path = self._normalize_model_path(model_path) or model_path
+
+            # Create inference engine
+            self.inference_engine = InferenceEngine(
+                normalized_model_path, self.db_manager, model_id
+            )
+
+            # Get confidence threshold
+            conf = self.conf_slider.value() / 100.0
+
+            # Get repository root
+            repo_root = self.config_manager.get_image_repository_path()
+            if not repo_root:
+                repo_root = str(Path(image_paths[0]).parent)
+
+            # Show progress bar
+            self.progress_bar.setVisible(True)
+            self.progress_bar.setMaximum(len(image_paths))
+            self._set_buttons_enabled(False)
+
+            # Create and start worker thread
+            self.worker = DetectionWorker(
+                self.inference_engine, image_paths, repo_root, conf
+            )
+            self.worker.progress.connect(self._on_progress)
+            self.worker.finished.connect(self._on_detection_finished)
+            self.worker.error.connect(self._on_detection_error)
+            self.worker.start()
+
+        except Exception as e:
+            logger.error(f"Error starting detection: {e}")
+            QMessageBox.critical(self, "Error", f"Failed to start detection:\n{str(e)}")
+            self._set_buttons_enabled(True)
+
+    def _on_progress(self, current: int, total: int, message: str):
+        """Handle progress update."""
+        self.progress_bar.setValue(current)
+        self.results_text.append(f"[{current}/{total}] {message}")
+
+    def _on_detection_finished(self, results: list):
+        """Handle detection completion."""
+        self.progress_bar.setVisible(False)
+        self._set_buttons_enabled(True)
+
+        # Calculate statistics
+        total_detections = sum(r["count"] for r in results)
+        successful = sum(1 for r in results if r.get("success", False))
+
+        summary = f"\n=== Detection Complete ===\n"
+        summary += f"Processed: {len(results)} images\n"
+        summary += f"Successful: {successful}\n"
+        summary += f"Total detections: {total_detections}\n"
+
+        self.results_text.append(summary)
+
+        QMessageBox.information(
+            self,
+            "Detection Complete",
+            f"Processed {len(results)} images\n{total_detections} objects detected",
+        )
+
+    def _on_detection_error(self, error_msg: str):
+        """Handle detection error."""
+        self.progress_bar.setVisible(False)
+        self._set_buttons_enabled(True)
+
+        self.results_text.append(f"\nERROR: {error_msg}")
+        QMessageBox.critical(self, "Detection Error", error_msg)
+
+    def _set_buttons_enabled(self, enabled: bool):
+        """Enable/disable action buttons."""
+        self.single_image_btn.setEnabled(enabled)
+        self.batch_btn.setEnabled(enabled)
+        self.model_combo.setEnabled(enabled)
+
+    def _discover_local_models(self) -> list:
+        """Scan the models directory for standalone .pt files."""
+        models_dir = self.config_manager.get_models_directory()
+        if not models_dir:
+            return []
+
+        models_path = Path(models_dir)
+        if not models_path.exists():
+            return []
+
+        try:
+            return sorted(
+                [p for p in models_path.rglob("*.pt") if p.is_file()],
+                key=lambda p: str(p).lower(),
+            )
+        except Exception as e:
+            logger.warning(f"Error discovering local models: {e}")
+            return []
+
+    def _normalize_model_path(self, path_value) -> str:
+        """Return a normalized absolute path string for comparison."""
+        if not path_value:
+            return ""
+        try:
+            return str(Path(path_value).resolve())
+        except Exception:
+            return str(path_value)
+
+    def _ensure_model_record(self, model_data: dict) -> Optional[int]:
+        """Ensure a database record exists for the selected model."""
+        model_path = model_data.get("path")
+        if not model_path:
+            return None
+
+        normalized_target = self._normalize_model_path(model_path)
+
+        try:
+            existing_models = self.db_manager.get_models()
+            for model in existing_models:
+                existing_path = model.get("model_path")
+                if not existing_path:
+                    continue
+                normalized_existing = self._normalize_model_path(existing_path)
+                if (
+                    normalized_existing == normalized_target
+                    or existing_path == model_path
+                ):
+                    return model["id"]
+
+            model_name = (
+                model_data.get("model_name") or Path(model_path).stem or "Custom Model"
+            )
+            model_version = (
+                model_data.get("model_version") or model_data.get("source") or "local"
+            )
+            base_model = model_data.get(
+                "base_model",
+                self.config_manager.get("models.default_base_model", "yolov8s-seg.pt"),
+            )
+
+            return self.db_manager.add_model(
+                model_name=model_name,
+                model_version=model_version,
+                model_path=normalized_target,
+                base_model=base_model,
+            )
+        except Exception as e:
+            logger.error(f"Failed to ensure model record for {model_path}: {e}")
+            return None
+
+    def refresh(self):
+        """Refresh the tab."""
+        self._load_models()
+        self.results_text.clear()

src/gui/tabs/results_tab.py

@@ -0,0 +1,439 @@
+"""
+Results tab for browsing stored detections and visualizing overlays.
+"""
+
+from pathlib import Path
+from typing import Dict, List, Optional
+
+from PySide6.QtWidgets import (
+    QWidget,
+    QVBoxLayout,
+    QHBoxLayout,
+    QLabel,
+    QGroupBox,
+    QPushButton,
+    QSplitter,
+    QTableWidget,
+    QTableWidgetItem,
+    QHeaderView,
+    QAbstractItemView,
+    QMessageBox,
+    QCheckBox,
+)
+from PySide6.QtCore import Qt
+
+from src.database.db_manager import DatabaseManager
+from src.utils.config_manager import ConfigManager
+from src.utils.logger import get_logger
+from src.utils.image import Image, ImageLoadError
+from src.gui.widgets import AnnotationCanvasWidget
+
+
+logger = get_logger(__name__)
+
+
+class ResultsTab(QWidget):
+    """Results tab showing detection history and preview overlays."""
+
+    def __init__(
+        self, db_manager: DatabaseManager, config_manager: ConfigManager, parent=None
+    ):
+        super().__init__(parent)
+        self.db_manager = db_manager
+        self.config_manager = config_manager
+
+        self.detection_summary: List[Dict] = []
+        self.current_selection: Optional[Dict] = None
+        self.current_image: Optional[Image] = None
+        self.current_detections: List[Dict] = []
+        self._image_path_cache: Dict[str, str] = {}
+
+        self._setup_ui()
+        self.refresh()
+
+    def _setup_ui(self):
+        """Setup user interface."""
+        layout = QVBoxLayout()
+
+        # Splitter for list + preview
+        splitter = QSplitter(Qt.Horizontal)
+
+        # Left pane: detection list
+        left_container = QWidget()
+        left_layout = QVBoxLayout()
+        left_layout.setContentsMargins(0, 0, 0, 0)
+
+        controls_layout = QHBoxLayout()
+        self.refresh_btn = QPushButton("Refresh")
+        self.refresh_btn.clicked.connect(self.refresh)
+        controls_layout.addWidget(self.refresh_btn)
+        controls_layout.addStretch()
+        left_layout.addLayout(controls_layout)
+
+        self.results_table = QTableWidget(0, 5)
+        self.results_table.setHorizontalHeaderLabels(
+            ["Image", "Model", "Detections", "Classes", "Last Updated"]
+        )
+        self.results_table.horizontalHeader().setSectionResizeMode(
+            0, QHeaderView.Stretch
+        )
+        self.results_table.horizontalHeader().setSectionResizeMode(
+            1, QHeaderView.Stretch
+        )
+        self.results_table.horizontalHeader().setSectionResizeMode(
+            2, QHeaderView.ResizeToContents
+        )
+        self.results_table.horizontalHeader().setSectionResizeMode(
+            3, QHeaderView.Stretch
+        )
+        self.results_table.horizontalHeader().setSectionResizeMode(
+            4, QHeaderView.ResizeToContents
+        )
+        self.results_table.setSelectionBehavior(QAbstractItemView.SelectRows)
+        self.results_table.setSelectionMode(QAbstractItemView.SingleSelection)
+        self.results_table.setEditTriggers(QAbstractItemView.NoEditTriggers)
+        self.results_table.itemSelectionChanged.connect(self._on_result_selected)
+
+        left_layout.addWidget(self.results_table)
+        left_container.setLayout(left_layout)
+
+        # Right pane: preview canvas and controls
+        right_container = QWidget()
+        right_layout = QVBoxLayout()
+        right_layout.setContentsMargins(0, 0, 0, 0)
+
+        preview_group = QGroupBox("Detection Preview")
+        preview_layout = QVBoxLayout()
+
+        self.preview_canvas = AnnotationCanvasWidget()
+        self.preview_canvas.set_polyline_enabled(False)
+        self.preview_canvas.set_show_bboxes(True)
+        preview_layout.addWidget(self.preview_canvas)
+
+        toggles_layout = QHBoxLayout()
+        self.show_masks_checkbox = QCheckBox("Show Masks")
+        self.show_masks_checkbox.setChecked(False)
+        self.show_masks_checkbox.stateChanged.connect(self._apply_detection_overlays)
+        self.show_bboxes_checkbox = QCheckBox("Show Bounding Boxes")
+        self.show_bboxes_checkbox.setChecked(True)
+        self.show_bboxes_checkbox.stateChanged.connect(self._toggle_bboxes)
+        self.show_confidence_checkbox = QCheckBox("Show Confidence")
+        self.show_confidence_checkbox.setChecked(False)
+        self.show_confidence_checkbox.stateChanged.connect(
+            self._apply_detection_overlays
+        )
+        toggles_layout.addWidget(self.show_masks_checkbox)
+        toggles_layout.addWidget(self.show_bboxes_checkbox)
+        toggles_layout.addWidget(self.show_confidence_checkbox)
+        toggles_layout.addStretch()
+        preview_layout.addLayout(toggles_layout)
+
+        self.summary_label = QLabel("Select a detection result to preview.")
+        self.summary_label.setWordWrap(True)
+        preview_layout.addWidget(self.summary_label)
+
+        preview_group.setLayout(preview_layout)
+        right_layout.addWidget(preview_group)
+        right_container.setLayout(right_layout)
+
+        splitter.addWidget(left_container)
+        splitter.addWidget(right_container)
+        splitter.setStretchFactor(0, 1)
+        splitter.setStretchFactor(1, 2)
+
+        layout.addWidget(splitter)
+        self.setLayout(layout)
+
+    def refresh(self):
+        """Refresh the detection list and preview."""
+        self._load_detection_summary()
+        self._populate_results_table()
+        self.current_selection = None
+        self.current_image = None
+        self.current_detections = []
+        self.preview_canvas.clear()
+        self.summary_label.setText("Select a detection result to preview.")
+
+    def _load_detection_summary(self):
+        """Load latest detection summaries grouped by image + model."""
+        try:
+            detections = self.db_manager.get_detections(limit=500)
+            summary_map: Dict[tuple, Dict] = {}
+
+            for det in detections:
+                key = (det["image_id"], det["model_id"])
+                metadata = det.get("metadata") or {}
+                entry = summary_map.setdefault(
+                    key,
+                    {
+                        "image_id": det["image_id"],
+                        "model_id": det["model_id"],
+                        "image_path": det.get("image_path"),
+                        "image_filename": det.get("image_filename")
+                        or det.get("image_path"),
+                        "model_name": det.get("model_name", ""),
+                        "model_version": det.get("model_version", ""),
+                        "last_detected": det.get("detected_at"),
+                        "count": 0,
+                        "classes": set(),
+                        "source_path": metadata.get("source_path"),
+                        "repository_root": metadata.get("repository_root"),
+                    },
+                )
+
+                entry["count"] += 1
+                if det.get("detected_at") and (
+                    not entry.get("last_detected")
+                    or str(det.get("detected_at")) > str(entry.get("last_detected"))
+                ):
+                    entry["last_detected"] = det.get("detected_at")
+                if det.get("class_name"):
+                    entry["classes"].add(det["class_name"])
+                if metadata.get("source_path") and not entry.get("source_path"):
+                    entry["source_path"] = metadata.get("source_path")
+                if metadata.get("repository_root") and not entry.get("repository_root"):
+                    entry["repository_root"] = metadata.get("repository_root")
+
+            self.detection_summary = sorted(
+                summary_map.values(),
+                key=lambda x: str(x.get("last_detected") or ""),
+                reverse=True,
+            )
+        except Exception as e:
+            logger.error(f"Failed to load detection summary: {e}")
+            QMessageBox.critical(
+                self,
+                "Error",
+                f"Failed to load detection results:\n{str(e)}",
+            )
+            self.detection_summary = []
+
+    def _populate_results_table(self):
+        """Populate the table widget with detection summaries."""
+        self.results_table.setRowCount(len(self.detection_summary))
+
+        for row, entry in enumerate(self.detection_summary):
+            model_label = f"{entry['model_name']} {entry['model_version']}".strip()
+            class_list = (
+                ", ".join(sorted(entry["classes"])) if entry["classes"] else "-"
+            )
+
+            items = [
+                QTableWidgetItem(entry.get("image_filename", "")),
+                QTableWidgetItem(model_label),
+                QTableWidgetItem(str(entry.get("count", 0))),
+                QTableWidgetItem(class_list),
+                QTableWidgetItem(str(entry.get("last_detected") or "")),
+            ]
+
+            for col, item in enumerate(items):
+                item.setData(Qt.UserRole, row)
+                self.results_table.setItem(row, col, item)
+
+        self.results_table.clearSelection()
+
+    def _on_result_selected(self):
+        """Handle selection changes in the detection table."""
+        selected_items = self.results_table.selectedItems()
+        if not selected_items:
+            return
+
+        row = selected_items[0].data(Qt.UserRole)
+        if row is None or row >= len(self.detection_summary):
+            return
+
+        entry = self.detection_summary[row]
+        if (
+            self.current_selection
+            and self.current_selection.get("image_id") == entry["image_id"]
+            and self.current_selection.get("model_id") == entry["model_id"]
+        ):
+            return
+
+        self.current_selection = entry
+
+        image_path = self._resolve_image_path(entry)
+        if not image_path:
+            QMessageBox.warning(
+                self,
+                "Image Not Found",
+                "Unable to locate the image file for this detection.",
+            )
+            return
+
+        try:
+            self.current_image = Image(image_path)
+            self.preview_canvas.load_image(self.current_image)
+        except ImageLoadError as e:
+            logger.error(f"Failed to load image '{image_path}': {e}")
+            QMessageBox.critical(
+                self,
+                "Image Error",
+                f"Failed to load image for preview:\n{str(e)}",
+            )
+            return
+
+        self._load_detections_for_selection(entry)
+        self._apply_detection_overlays()
+        self._update_summary_label(entry)
+
+    def _load_detections_for_selection(self, entry: Dict):
+        """Load detection records for the selected image/model pair."""
+        self.current_detections = []
+        if not entry:
+            return
+
+        try:
+            filters = {"image_id": entry["image_id"], "model_id": entry["model_id"]}
+            self.current_detections = self.db_manager.get_detections(filters)
+        except Exception as e:
+            logger.error(f"Failed to load detections for preview: {e}")
+            QMessageBox.critical(
+                self,
+                "Error",
+                f"Failed to load detections for this image:\n{str(e)}",
+            )
+            self.current_detections = []
+
+    def _apply_detection_overlays(self):
+        """Draw detections onto the preview canvas based on current toggles."""
+        self.preview_canvas.clear_annotations()
+        self.preview_canvas.set_show_bboxes(self.show_bboxes_checkbox.isChecked())
+
+        if not self.current_detections or not self.current_image:
+            return
+
+        for det in self.current_detections:
+            color = self._get_class_color(det.get("class_name"))
+
+            if self.show_masks_checkbox.isChecked() and det.get("segmentation_mask"):
+                mask_points = self._convert_mask(det["segmentation_mask"])
+                if mask_points:
+                    self.preview_canvas.draw_saved_polyline(mask_points, color)
+
+            bbox = [
+                det.get("x_min"),
+                det.get("y_min"),
+                det.get("x_max"),
+                det.get("y_max"),
+            ]
+            if all(v is not None for v in bbox):
+                label = None
+                if self.show_confidence_checkbox.isChecked():
+                    confidence = det.get("confidence")
+                    if confidence is not None:
+                        label = f"{confidence:.2f}"
+                self.preview_canvas.draw_saved_bbox(bbox, color, label=label)
+
+    def _convert_mask(self, mask_points: List[List[float]]) -> List[List[float]]:
+        """Convert stored [x, y] masks to [y, x] format for the canvas."""
+        converted = []
+        for point in mask_points:
+            if len(point) >= 2:
+                x, y = point[0], point[1]
+                converted.append([y, x])
+        return converted
+
+    def _toggle_bboxes(self):
+        """Update bounding box visibility on the canvas."""
+        self.preview_canvas.set_show_bboxes(self.show_bboxes_checkbox.isChecked())
+        # Re-render to respect show/hide when toggled
+        self._apply_detection_overlays()
+
+    def _update_summary_label(self, entry: Dict):
+        """Display textual summary for the selected detection run."""
+        classes = ", ".join(sorted(entry.get("classes", []))) or "-"
+        summary_text = (
+            f"Image: {entry.get('image_filename', 'unknown')}\n"
+            f"Model: {entry.get('model_name', '')} {entry.get('model_version', '')}\n"
+            f"Detections: {entry.get('count', 0)}\n"
+            f"Classes: {classes}\n"
+            f"Last Updated: {entry.get('last_detected', 'n/a')}"
+        )
+        self.summary_label.setText(summary_text)
+
+    def _resolve_image_path(self, entry: Dict) -> Optional[str]:
+        """Resolve an image path using metadata, cache, and repository hints."""
+        relative_path = entry.get("image_path") if entry else None
+        cache_key = relative_path or entry.get("source_path")
+        if cache_key and cache_key in self._image_path_cache:
+            cached = Path(self._image_path_cache[cache_key])
+            if cached.exists():
+                return self._image_path_cache[cache_key]
+            del self._image_path_cache[cache_key]
+
+        candidates = []
+        source_path = entry.get("source_path") if entry else None
+        if source_path:
+            candidates.append(Path(source_path))
+
+        repo_roots = []
+        if entry.get("repository_root"):
+            repo_roots.append(entry["repository_root"])
+        config_repo = self.config_manager.get_image_repository_path()
+        if config_repo:
+            repo_roots.append(config_repo)
+
+        for root in repo_roots:
+            if relative_path:
+                candidates.append(Path(root) / relative_path)
+
+        if relative_path:
+            candidates.append(Path(relative_path))
+
+        for candidate in candidates:
+            try:
+                if candidate and candidate.exists():
+                    resolved = str(candidate.resolve())
+                    if cache_key:
+                        self._image_path_cache[cache_key] = resolved
+                    return resolved
+            except Exception:
+                continue
+
+        # Fallback: search by filename in known roots
+        filename = Path(relative_path).name if relative_path else None
+        if filename:
+            search_roots = [Path(root) for root in repo_roots if root]
+            if not search_roots:
+                search_roots = [Path("data")]
+            match = self._search_in_roots(filename, search_roots)
+            if match:
+                resolved = str(match.resolve())
+                if cache_key:
+                    self._image_path_cache[cache_key] = resolved
+                return resolved
+
+        return None
+
+    def _search_in_roots(self, filename: str, roots: List[Path]) -> Optional[Path]:
+        """Search for a file name within a list of root directories."""
+        for root in roots:
+            try:
+                if not root.exists():
+                    continue
+                for candidate in root.rglob(filename):
+                    return candidate
+            except Exception as e:
+                logger.debug(f"Error searching for {filename} in {root}: {e}")
+        return None
+
+    def _get_class_color(self, class_name: Optional[str]) -> str:
+        """Return consistent color hex for a class name."""
+        if not class_name:
+            return "#FF6B6B"
+
+        color_map = self.config_manager.get_bbox_colors()
+        if class_name in color_map:
+            return color_map[class_name]
+
+        # Deterministic fallback color based on hash
+        palette = [
+            "#FF6B6B",
+            "#4ECDC4",
+            "#FFD166",
+            "#1D3557",
+            "#F4A261",
+            "#E76F51",
+        ]
+        return palette[hash(class_name) % len(palette)]

src/gui/tabs/validation_tab.py

@@ -0,0 +1,46 @@
+"""
+Validation tab for the microscopy object detection application.
+"""
+
+from PySide6.QtWidgets import QWidget, QVBoxLayout, QLabel, QGroupBox
+
+from src.database.db_manager import DatabaseManager
+from src.utils.config_manager import ConfigManager
+
+
+class ValidationTab(QWidget):
+    """Validation tab placeholder."""
+
+    def __init__(
+        self, db_manager: DatabaseManager, config_manager: ConfigManager, parent=None
+    ):
+        super().__init__(parent)
+        self.db_manager = db_manager
+        self.config_manager = config_manager
+
+        self._setup_ui()
+
+    def _setup_ui(self):
+        """Setup user interface."""
+        layout = QVBoxLayout()
+
+        group = QGroupBox("Validation")
+        group_layout = QVBoxLayout()
+        label = QLabel(
+            "Validation functionality will be implemented here.\n\n"
+            "Features:\n"
+            "- Model validation\n"
+            "- Metrics visualization\n"
+            "- Confusion matrix\n"
+            "- Precision-Recall curves"
+        )
+        group_layout.addWidget(label)
+        group.setLayout(group_layout)
+
+        layout.addWidget(group)
+        layout.addStretch()
+        self.setLayout(layout)
+
+    def refresh(self):
+        """Refresh the tab."""
+        pass

src/gui/widgets/__init__.py

@@ -0,0 +1,7 @@
+"""GUI widgets for the microscopy object detection application."""
+
+from src.gui.widgets.image_display_widget import ImageDisplayWidget
+from src.gui.widgets.annotation_canvas_widget import AnnotationCanvasWidget
+from src.gui.widgets.annotation_tools_widget import AnnotationToolsWidget
+
+__all__ = ["ImageDisplayWidget", "AnnotationCanvasWidget", "AnnotationToolsWidget"]

src/gui/widgets/annotation_canvas_widget.py

@@ -0,0 +1,931 @@
+"""
+Annotation canvas widget for drawing annotations on images.
+Currently supports polyline drawing tool with color selection for manual annotation.
+"""
+
+import numpy as np
+import math
+
+from PySide6.QtWidgets import QWidget, QVBoxLayout, QLabel, QScrollArea
+from PySide6.QtGui import (
+    QPixmap,
+    QImage,
+    QPainter,
+    QPen,
+    QColor,
+    QKeyEvent,
+    QMouseEvent,
+    QPaintEvent,
+    QPolygonF,
+)
+from PySide6.QtCore import Qt, QEvent, Signal, QPoint, QPointF, QRect
+from typing import Any, Dict, List, Optional, Tuple
+
+from src.utils.image import Image, ImageLoadError
+from src.utils.logger import get_logger
+
+logger = get_logger(__name__)
+
+
+def perpendicular_distance(
+    point: Tuple[float, float],
+    start: Tuple[float, float],
+    end: Tuple[float, float],
+) -> float:
+    """Perpendicular distance from `point` to the line defined by `start`->`end`."""
+    (x, y), (x1, y1), (x2, y2) = point, start, end
+    dx = x2 - x1
+    dy = y2 - y1
+    if dx == 0.0 and dy == 0.0:
+        return math.hypot(x - x1, y - y1)
+    num = abs(dy * x - dx * y + x2 * y1 - y2 * x1)
+    den = math.hypot(dx, dy)
+    return num / den
+
+
+def rdp(points: List[Tuple[float, float]], epsilon: float) -> List[Tuple[float, float]]:
+    """
+    Recursive Ramer-Douglas-Peucker (RDP) polyline simplification.
+
+    Args:
+        points: List of (x, y) points.
+        epsilon: Maximum allowed perpendicular distance in pixels.
+
+    Returns:
+        Simplified list of (x, y) points including first and last.
+    """
+    if len(points) <= 2:
+        return list(points)
+
+    start = points[0]
+    end = points[-1]
+    max_dist = -1.0
+    index = -1
+
+    for i in range(1, len(points) - 1):
+        d = perpendicular_distance(points[i], start, end)
+        if d > max_dist:
+            max_dist = d
+            index = i
+
+    if max_dist > epsilon:
+        # Recursive split
+        left = rdp(points[: index + 1], epsilon)
+        right = rdp(points[index:], epsilon)
+        # Concatenate but avoid duplicate at split point
+        return left[:-1] + right
+
+    # Keep only start and end
+    return [start, end]
+
+
+def simplify_polyline(
+    points: List[Tuple[float, float]], epsilon: float
+) -> List[Tuple[float, float]]:
+    """
+    Simplify a polyline with RDP while preserving closure semantics.
+
+    If the polyline is closed (first == last), the duplicate last point is removed
+    before simplification and then re-added after simplification.
+    """
+    if not points:
+        return []
+
+    pts = [(float(x), float(y)) for x, y in points]
+    closed = False
+
+    if len(pts) >= 2 and pts[0] == pts[-1]:
+        closed = True
+        pts = pts[:-1]  # remove duplicate last for simplification
+
+    if len(pts) <= 2:
+        simplified = list(pts)
+    else:
+        simplified = rdp(pts, epsilon)
+
+    if closed and simplified:
+        if simplified[0] != simplified[-1]:
+            simplified.append(simplified[0])
+
+    return simplified
+
+
+class AnnotationCanvasWidget(QWidget):
+    """
+    Widget for displaying images and drawing annotations with zoom and drawing tools.
+
+    Features:
+    - Display images with zoom functionality
+    - Polyline tool for drawing annotations
+    - Configurable pen color and width
+    - Mouse-based drawing interface
+    - Zoom in/out with mouse wheel and keyboard
+
+    Signals:
+        zoom_changed: Emitted when zoom level changes (float zoom_scale)
+        annotation_drawn: Emitted when a new stroke is completed (list of points)
+    """
+
+    zoom_changed = Signal(float)
+    annotation_drawn = Signal(list)  # List of (x, y) points in normalized coordinates
+    # Emitted when the user selects an existing polyline on the canvas.
+    # Carries the associated annotation_id (int) or None if selection is cleared
+    annotation_selected = Signal(object)
+
+    def __init__(self, parent=None):
+        """Initialize the annotation canvas widget."""
+        super().__init__(parent)
+
+        self.current_image = None
+        self.original_pixmap = None
+        self.annotation_pixmap = None  # Overlay for annotations
+        self.zoom_scale = 1.0
+        self.zoom_min = 0.1
+        self.zoom_max = 10.0
+        self.zoom_step = 0.1
+        self.zoom_wheel_step = 0.15
+
+        # Drawing / interaction state
+        self.is_drawing = False
+        self.polyline_enabled = False
+        self.polyline_pen_color = QColor(255, 0, 0, 128)  # Default red with 50% alpha
+        self.polyline_pen_width = 3
+        self.show_bboxes: bool = True  # Control visibility of bounding boxes
+
+        # Current stroke and stored polylines (in image coordinates, pixel units)
+        self.current_stroke: List[Tuple[float, float]] = []
+        self.polylines: List[List[Tuple[float, float]]] = []
+        self.stroke_meta: List[Dict[str, Any]] = []  # per-polyline style (color, width)
+        # Optional DB annotation_id for each stored polyline (None for temporary / unsaved)
+        self.polyline_annotation_ids: List[Optional[int]] = []
+        # Indices in self.polylines of the currently selected polylines (multi-select)
+        self.selected_polyline_indices: List[int] = []
+
+        # Stored bounding boxes in normalized coordinates (x_min, y_min, x_max, y_max)
+        self.bboxes: List[List[float]] = []
+        self.bbox_meta: List[Dict[str, Any]] = []  # per-bbox style (color, width)
+
+        # Legacy collection of strokes in normalized coordinates (kept for API compatibility)
+        self.all_strokes: List[dict] = []
+
+        # RDP simplification parameters (in pixels)
+        self.simplify_on_finish: bool = True
+        self.simplify_epsilon: float = 2.0
+        self.sample_threshold: float = 2.0  # minimum movement to sample a new point
+
+        self._setup_ui()
+
+    def _setup_ui(self):
+        """Setup user interface."""
+        layout = QVBoxLayout()
+        layout.setContentsMargins(0, 0, 0, 0)
+
+        # Scroll area for canvas
+        self.scroll_area = QScrollArea()
+        self.scroll_area.setWidgetResizable(True)
+        self.scroll_area.setMinimumHeight(400)
+
+        self.canvas_label = QLabel("No image loaded")
+        self.canvas_label.setAlignment(Qt.AlignCenter)
+        self.canvas_label.setStyleSheet(
+            "QLabel { background-color: #2b2b2b; color: #888; }"
+        )
+        self.canvas_label.setScaledContents(False)
+        self.canvas_label.setMouseTracking(True)
+
+        self.scroll_area.setWidget(self.canvas_label)
+        self.scroll_area.viewport().installEventFilter(self)
+
+        layout.addWidget(self.scroll_area)
+        self.setLayout(layout)
+
+        self.setFocusPolicy(Qt.StrongFocus)
+
+    def load_image(self, image: Image):
+        """
+        Load and display an image.
+
+        Args:
+            image: Image object to display
+        """
+        self.current_image = image
+        self.zoom_scale = 1.0
+        self.clear_annotations()
+        self._display_image()
+        logger.debug(
+            f"Loaded image into annotation canvas: {image.width}x{image.height}"
+        )
+
+    def clear(self):
+        """Clear the displayed image and all annotations."""
+        self.current_image = None
+        self.original_pixmap = None
+        self.annotation_pixmap = None
+        self.zoom_scale = 1.0
+        self.clear_annotations()
+        self.canvas_label.setText("No image loaded")
+        self.canvas_label.setPixmap(QPixmap())
+
+    def clear_annotations(self):
+        """Clear all drawn annotations."""
+        self.all_strokes = []
+        self.current_stroke = []
+        self.polylines = []
+        self.stroke_meta = []
+        self.polyline_annotation_ids = []
+        self.selected_polyline_indices = []
+        self.bboxes = []
+        self.bbox_meta = []
+        self.is_drawing = False
+        if self.annotation_pixmap:
+            self.annotation_pixmap.fill(Qt.transparent)
+            self._update_display()
+
+    def _display_image(self):
+        """Display the current image in the canvas."""
+        if self.current_image is None:
+            return
+
+        try:
+            # Get image data in a format compatible with Qt
+            if self.current_image.channels in (3, 4):
+                image_data = self.current_image.get_rgb()
+                height, width = image_data.shape[:2]
+            else:
+                image_data = self.current_image.get_grayscale()
+                height, width = image_data.shape
+
+            image_data = np.ascontiguousarray(image_data)
+            bytes_per_line = image_data.strides[0]
+
+            qimage = QImage(
+                image_data.data,
+                width,
+                height,
+                bytes_per_line,
+                self.current_image.qtimage_format,
+            ).copy()  # Copy so Qt owns the buffer even after numpy array goes out of scope
+
+            self.original_pixmap = QPixmap.fromImage(qimage)
+
+            # Create transparent overlay for annotations
+            self.annotation_pixmap = QPixmap(self.original_pixmap.size())
+            self.annotation_pixmap.fill(Qt.transparent)
+
+            self._apply_zoom()
+
+        except Exception as e:
+            logger.error(f"Error displaying image: {e}")
+            raise ImageLoadError(f"Failed to display image: {str(e)}")
+
+    def _apply_zoom(self):
+        """Apply current zoom level to the displayed image."""
+        if self.original_pixmap is None:
+            return
+
+        scaled_width = int(self.original_pixmap.width() * self.zoom_scale)
+        scaled_height = int(self.original_pixmap.height() * self.zoom_scale)
+
+        # Scale both image and annotations
+        scaled_image = self.original_pixmap.scaled(
+            scaled_width,
+            scaled_height,
+            Qt.KeepAspectRatio,
+            (
+                Qt.SmoothTransformation
+                if self.zoom_scale >= 1.0
+                else Qt.FastTransformation
+            ),
+        )
+
+        scaled_annotations = self.annotation_pixmap.scaled(
+            scaled_width,
+            scaled_height,
+            Qt.KeepAspectRatio,
+            (
+                Qt.SmoothTransformation
+                if self.zoom_scale >= 1.0
+                else Qt.FastTransformation
+            ),
+        )
+
+        # Composite image and annotations
+        combined = QPixmap(scaled_image.size())
+        painter = QPainter(combined)
+        painter.drawPixmap(0, 0, scaled_image)
+        painter.drawPixmap(0, 0, scaled_annotations)
+        painter.end()
+
+        self.canvas_label.setPixmap(combined)
+        self.canvas_label.setScaledContents(False)
+        self.canvas_label.adjustSize()
+
+        self.zoom_changed.emit(self.zoom_scale)
+
+    def _update_display(self):
+        """Update display after drawing."""
+        self._apply_zoom()
+
+    def set_polyline_enabled(self, enabled: bool):
+        """Enable or disable polyline tool."""
+        self.polyline_enabled = enabled
+        if enabled:
+            self.canvas_label.setCursor(Qt.CrossCursor)
+        else:
+            self.canvas_label.setCursor(Qt.ArrowCursor)
+
+    def set_polyline_pen_color(self, color: QColor):
+        """Set polyline pen color."""
+        self.polyline_pen_color = color
+
+    def set_polyline_pen_width(self, width: int):
+        """Set polyline pen width."""
+        self.polyline_pen_width = max(1, width)
+
+    def get_zoom_percentage(self) -> int:
+        """Get current zoom level as percentage."""
+        return int(self.zoom_scale * 100)
+
+    def zoom_in(self):
+        """Zoom in on the image."""
+        if self.original_pixmap is None:
+            return
+        new_scale = self.zoom_scale + self.zoom_step
+        if new_scale <= self.zoom_max:
+            self.zoom_scale = new_scale
+            self._apply_zoom()
+
+    def zoom_out(self):
+        """Zoom out from the image."""
+        if self.original_pixmap is None:
+            return
+        new_scale = self.zoom_scale - self.zoom_step
+        if new_scale >= self.zoom_min:
+            self.zoom_scale = new_scale
+            self._apply_zoom()
+
+    def reset_zoom(self):
+        """Reset zoom to 100%."""
+        if self.original_pixmap is None:
+            return
+        self.zoom_scale = 1.0
+        self._apply_zoom()
+
+    def _canvas_to_image_coords(self, pos: QPoint) -> Optional[Tuple[int, int]]:
+        """Convert canvas coordinates to image coordinates, accounting for zoom and centering."""
+        if self.original_pixmap is None or self.canvas_label.pixmap() is None:
+            return None
+
+        # Get the displayed pixmap size (after zoom)
+        displayed_pixmap = self.canvas_label.pixmap()
+        displayed_width = displayed_pixmap.width()
+        displayed_height = displayed_pixmap.height()
+
+        # Calculate offset due to label centering (label might be larger than pixmap)
+        label_width = self.canvas_label.width()
+        label_height = self.canvas_label.height()
+        offset_x = max(0, (label_width - displayed_width) // 2)
+        offset_y = max(0, (label_height - displayed_height) // 2)
+
+        # Adjust position for offset and convert to image coordinates
+        x = (pos.x() - offset_x) / self.zoom_scale
+        y = (pos.y() - offset_y) / self.zoom_scale
+
+        # Check bounds
+        if (
+            0 <= x < self.original_pixmap.width()
+            and 0 <= y < self.original_pixmap.height()
+        ):
+            return (int(x), int(y))
+        return None
+
+    def _find_polyline_at(
+        self, img_x: float, img_y: float, threshold_px: float = 5.0
+    ) -> Optional[int]:
+        """
+        Find index of polyline whose geometry is within threshold_px of (img_x, img_y).
+        Returns the index in self.polylines, or None if none is close enough.
+        """
+        best_index: Optional[int] = None
+        best_dist: float = float("inf")
+
+        for idx, polyline in enumerate(self.polylines):
+            if len(polyline) < 2:
+                continue
+
+            # Quick bounding-box check to skip obviously distant polylines
+            xs = [p[0] for p in polyline]
+            ys = [p[1] for p in polyline]
+            if img_x < min(xs) - threshold_px or img_x > max(xs) + threshold_px:
+                continue
+            if img_y < min(ys) - threshold_px or img_y > max(ys) + threshold_px:
+                continue
+
+            # Precise distance to all segments
+            for (x1, y1), (x2, y2) in zip(polyline[:-1], polyline[1:]):
+                d = perpendicular_distance(
+                    (img_x, img_y), (float(x1), float(y1)), (float(x2), float(y2))
+                )
+                if d < best_dist:
+                    best_dist = d
+                    best_index = idx
+
+        if best_index is not None and best_dist <= threshold_px:
+            return best_index
+        return None
+
+    def _image_to_normalized_coords(self, x: int, y: int) -> Tuple[float, float]:
+        """Convert image coordinates to normalized coordinates (0-1)."""
+        if self.original_pixmap is None:
+            return (0.0, 0.0)
+
+        norm_x = x / self.original_pixmap.width()
+        norm_y = y / self.original_pixmap.height()
+        return (norm_x, norm_y)
+
+    def _add_polyline(
+        self,
+        img_points: List[Tuple[float, float]],
+        color: QColor,
+        width: int,
+        annotation_id: Optional[int] = None,
+    ):
+        """Store a polyline in image coordinates and redraw annotations."""
+        if not img_points or len(img_points) < 2:
+            return
+
+        # Ensure all points are tuples of floats
+        normalized_points = [(float(x), float(y)) for x, y in img_points]
+        self.polylines.append(normalized_points)
+        self.stroke_meta.append({"color": QColor(color), "width": int(width)})
+        self.polyline_annotation_ids.append(annotation_id)
+
+        self._redraw_annotations()
+
+    def _redraw_annotations(self):
+        """Redraw all stored polylines and (optionally) bounding boxes onto the annotation pixmap."""
+        if self.annotation_pixmap is None:
+            return
+
+        # Clear existing overlay
+        self.annotation_pixmap.fill(Qt.transparent)
+
+        painter = QPainter(self.annotation_pixmap)
+
+        # Draw polylines
+        for idx, (polyline, meta) in enumerate(zip(self.polylines, self.stroke_meta)):
+            pen_color: QColor = meta.get("color", self.polyline_pen_color)
+            width: int = meta.get("width", self.polyline_pen_width)
+
+            if idx in self.selected_polyline_indices:
+                # Highlight selected polylines in a distinct color / width
+                highlight_color = QColor(255, 255, 0, 200)  # yellow, semi-opaque
+                pen = QPen(
+                    highlight_color,
+                    width + 1,
+                    Qt.SolidLine,
+                    Qt.RoundCap,
+                    Qt.RoundJoin,
+                )
+            else:
+                pen = QPen(
+                    pen_color,
+                    width,
+                    Qt.SolidLine,
+                    Qt.RoundCap,
+                    Qt.RoundJoin,
+                )
+
+            painter.setPen(pen)
+            # Use QPolygonF for efficient polygon rendering (single call vs N-1 calls)
+            # drawPolygon() automatically closes the shape, ensuring proper visual closure
+            polygon = QPolygonF([QPointF(x, y) for x, y in polyline])
+            painter.drawPolygon(polygon)
+
+        # Draw bounding boxes (dashed) if enabled
+        if self.show_bboxes and self.original_pixmap is not None and self.bboxes:
+            img_width = float(self.original_pixmap.width())
+            img_height = float(self.original_pixmap.height())
+
+            for bbox, meta in zip(self.bboxes, self.bbox_meta):
+                if len(bbox) != 4:
+                    continue
+
+                x_min_norm, y_min_norm, x_max_norm, y_max_norm = bbox
+                x_min = int(x_min_norm * img_width)
+                y_min = int(y_min_norm * img_height)
+                x_max = int(x_max_norm * img_width)
+                y_max = int(y_max_norm * img_height)
+
+                rect_width = x_max - x_min
+                rect_height = y_max - y_min
+
+                pen_color: QColor = meta.get("color", QColor(255, 0, 0, 128))
+                width: int = meta.get("width", self.polyline_pen_width)
+                pen = QPen(
+                    pen_color,
+                    width,
+                    Qt.DashLine,
+                    Qt.SquareCap,
+                    Qt.MiterJoin,
+                )
+                painter.setPen(pen)
+                painter.drawRect(x_min, y_min, rect_width, rect_height)
+
+                label_text = meta.get("label")
+                if label_text:
+                    painter.save()
+                    font = painter.font()
+                    font.setPointSizeF(max(10.0, width + 4))
+                    painter.setFont(font)
+                    metrics = painter.fontMetrics()
+                    text_width = metrics.horizontalAdvance(label_text)
+                    text_height = metrics.height()
+                    padding = 4
+                    bg_width = text_width + padding * 2
+                    bg_height = text_height + padding * 2
+                    canvas_width = self.original_pixmap.width()
+                    canvas_height = self.original_pixmap.height()
+                    bg_x = max(0, min(x_min, canvas_width - bg_width))
+                    bg_y = y_min - bg_height
+                    if bg_y < 0:
+                        bg_y = min(y_min, canvas_height - bg_height)
+                        bg_y = max(0, bg_y)
+                    background_rect = QRect(bg_x, bg_y, bg_width, bg_height)
+                    background_color = QColor(pen_color)
+                    background_color.setAlpha(220)
+                    painter.fillRect(background_rect, background_color)
+                    text_color = QColor(0, 0, 0)
+                    if background_color.lightness() < 128:
+                        text_color = QColor(255, 255, 255)
+                    painter.setPen(text_color)
+                    painter.drawText(
+                        background_rect.adjusted(padding, padding, -padding, -padding),
+                        Qt.AlignLeft | Qt.AlignVCenter,
+                        label_text,
+                    )
+                    painter.restore()
+
+        painter.end()
+
+        self._update_display()
+
+    def mousePressEvent(self, event: QMouseEvent):
+        """Handle mouse press events for drawing and selecting polylines."""
+        if self.annotation_pixmap is None:
+            super().mousePressEvent(event)
+            return
+
+        # Map click to image coordinates
+        label_pos = self.canvas_label.mapFromGlobal(event.globalPos())
+        img_coords = self._canvas_to_image_coords(label_pos)
+
+        # Left button + drawing tool enabled -> start a new stroke
+        if event.button() == Qt.LeftButton and self.polyline_enabled:
+            if img_coords:
+                self.is_drawing = True
+                self.current_stroke = [(float(img_coords[0]), float(img_coords[1]))]
+            return
+
+        # Left button + drawing tool disabled -> attempt selection of existing polyline
+        if event.button() == Qt.LeftButton and not self.polyline_enabled:
+            if img_coords:
+                idx = self._find_polyline_at(float(img_coords[0]), float(img_coords[1]))
+                if idx is not None:
+                    if event.modifiers() & Qt.ShiftModifier:
+                        # Multi-select mode: add to current selection (if not already selected)
+                        if idx not in self.selected_polyline_indices:
+                            self.selected_polyline_indices.append(idx)
+                    else:
+                        # Single-select mode: replace current selection
+                        self.selected_polyline_indices = [idx]
+
+                    # Build list of selected annotation IDs (ignore None entries)
+                    selected_ids: List[int] = []
+                    for sel_idx in self.selected_polyline_indices:
+                        if 0 <= sel_idx < len(self.polyline_annotation_ids):
+                            ann_id = self.polyline_annotation_ids[sel_idx]
+                            if isinstance(ann_id, int):
+                                selected_ids.append(ann_id)
+
+                    if selected_ids:
+                        self.annotation_selected.emit(selected_ids)
+                    else:
+                        # No valid DB-backed annotations in selection
+                        self.annotation_selected.emit(None)
+                else:
+                    # Clicked on empty space -> clear selection
+                    self.selected_polyline_indices = []
+                    self.annotation_selected.emit(None)
+
+                self._redraw_annotations()
+            return
+
+        # Fallback for other buttons / cases
+        super().mousePressEvent(event)
+
+    def mouseMoveEvent(self, event: QMouseEvent):
+        """Handle mouse move events for drawing."""
+        if (
+            not self.is_drawing
+            or not self.polyline_enabled
+            or self.annotation_pixmap is None
+        ):
+            super().mouseMoveEvent(event)
+            return
+
+        # Get accurate position using global coordinates
+        label_pos = self.canvas_label.mapFromGlobal(event.globalPos())
+        img_coords = self._canvas_to_image_coords(label_pos)
+
+        if img_coords and len(self.current_stroke) > 0:
+            last_point = self.current_stroke[-1]
+            dx = img_coords[0] - last_point[0]
+            dy = img_coords[1] - last_point[1]
+
+            # Only sample a new point if we moved enough pixels
+            if math.hypot(dx, dy) < self.sample_threshold:
+                return
+
+            # Draw line from last point to current point for interactive feedback
+            painter = QPainter(self.annotation_pixmap)
+            pen = QPen(
+                self.polyline_pen_color,
+                self.polyline_pen_width,
+                Qt.SolidLine,
+                Qt.RoundCap,
+                Qt.RoundJoin,
+            )
+            painter.setPen(pen)
+            painter.drawLine(
+                int(last_point[0]),
+                int(last_point[1]),
+                int(img_coords[0]),
+                int(img_coords[1]),
+            )
+            painter.end()
+
+            self.current_stroke.append((float(img_coords[0]), float(img_coords[1])))
+            self._update_display()
+
+    def mouseReleaseEvent(self, event: QMouseEvent):
+        """Handle mouse release events to complete a stroke."""
+        if not self.is_drawing or event.button() != Qt.LeftButton:
+            super().mouseReleaseEvent(event)
+            return
+
+        self.is_drawing = False
+
+        if len(self.current_stroke) > 1 and self.original_pixmap is not None:
+            # Ensure the stroke is closed by connecting end -> start
+            raw_points = list(self.current_stroke)
+            if raw_points[0] != raw_points[-1]:
+                raw_points.append(raw_points[0])
+
+            # Optional RDP simplification (in image pixel space)
+            if self.simplify_on_finish:
+                simplified = simplify_polyline(raw_points, self.simplify_epsilon)
+            else:
+                simplified = raw_points
+
+            if len(simplified) >= 2:
+                # Store polyline and redraw all annotations
+                self._add_polyline(
+                    simplified, self.polyline_pen_color, self.polyline_pen_width
+                )
+
+                # Convert to normalized coordinates for metadata + signal
+                normalized_stroke = [
+                    self._image_to_normalized_coords(int(x), int(y))
+                    for (x, y) in simplified
+                ]
+                self.all_strokes.append(
+                    {
+                        "points": normalized_stroke,
+                        "color": self.polyline_pen_color.name(),
+                        "alpha": self.polyline_pen_color.alpha(),
+                        "width": self.polyline_pen_width,
+                    }
+                )
+
+                # Emit signal with normalized coordinates
+                self.annotation_drawn.emit(normalized_stroke)
+                logger.debug(
+                    f"Completed stroke with {len(simplified)} points "
+                    f"(normalized len={len(normalized_stroke)})"
+                )
+
+        self.current_stroke = []
+
+    def get_all_strokes(self) -> List[dict]:
+        """Get all drawn strokes with metadata."""
+        return self.all_strokes
+
+    def get_annotation_parameters(self) -> Optional[List[Dict[str, Any]]]:
+        """
+        Get all annotation parameters including bounding box and polyline.
+
+        Returns:
+            List of dictionaries, each containing:
+            - 'bbox': [x_min, y_min, x_max, y_max] in normalized image coordinates
+            - 'polyline': List of [y_norm, x_norm] points describing the polygon
+        """
+        if self.original_pixmap is None or not self.polylines:
+            return None
+
+        img_width = float(self.original_pixmap.width())
+        img_height = float(self.original_pixmap.height())
+
+        params: List[Dict[str, Any]] = []
+
+        for idx, polyline in enumerate(self.polylines):
+            if len(polyline) < 2:
+                continue
+
+            xs = [p[0] for p in polyline]
+            ys = [p[1] for p in polyline]
+
+            x_min_norm = min(xs) / img_width
+            x_max_norm = max(xs) / img_width
+            y_min_norm = min(ys) / img_height
+            y_max_norm = max(ys) / img_height
+
+            # Store polyline as [y_norm, x_norm] to match DB convention and
+            # the expectations of draw_saved_polyline().
+            normalized_polyline = [
+                [y / img_height, x / img_width] for (x, y) in polyline
+            ]
+
+            logger.debug(
+                f"Polyline {idx}: {len(polyline)} points, "
+                f"bbox=({x_min_norm:.3f}, {y_min_norm:.3f})-({x_max_norm:.3f}, {y_max_norm:.3f})"
+            )
+
+            params.append(
+                {
+                    "bbox": [x_min_norm, y_min_norm, x_max_norm, y_max_norm],
+                    "polyline": normalized_polyline,
+                }
+            )
+
+        return params or None
+
+    def draw_saved_polyline(
+        self,
+        polyline: List[List[float]],
+        color: str,
+        width: int = 3,
+        annotation_id: Optional[int] = None,
+    ):
+        """
+        Draw a polyline from database coordinates onto the annotation canvas.
+
+        Args:
+            polyline: List of [x, y] coordinate pairs in normalized coordinates (0-1)
+            color: Color hex string (e.g., '#FF0000')
+            width: Line width in pixels
+        """
+        if not self.annotation_pixmap or not self.original_pixmap:
+            logger.warning("Cannot draw polyline: no image loaded")
+            return
+
+        if len(polyline) < 2:
+            logger.warning("Polyline has less than 2 points, cannot draw")
+            return
+
+        # Convert normalized coordinates to image coordinates
+        # Polyline is stored as [[y_norm, x_norm], ...] (row_norm, col_norm format)
+        img_width = self.original_pixmap.width()
+        img_height = self.original_pixmap.height()
+
+        logger.debug(f"Loading polyline with {len(polyline)} points")
+        logger.debug(f"  Image size: {img_width}x{img_height}")
+        logger.debug(f"  First 3 normalized points from DB: {polyline[:3]}")
+
+        img_coords: List[Tuple[float, float]] = []
+        for y_norm, x_norm in polyline:
+            x = float(x_norm * img_width)
+            y = float(y_norm * img_height)
+            img_coords.append((x, y))
+
+        logger.debug(f"  First 3 pixel coords: {img_coords[:3]}")
+
+        # Store and redraw using common pipeline
+        pen_color = QColor(color)
+        pen_color.setAlpha(128)  # Add semi-transparency
+        self._add_polyline(img_coords, pen_color, width, annotation_id=annotation_id)
+
+        # Store in all_strokes for consistency (uses normalized coordinates)
+        self.all_strokes.append(
+            {"points": polyline, "color": color, "alpha": 128, "width": width}
+        )
+
+        logger.debug(
+            f"Drew saved polyline with {len(polyline)} points in color {color}"
+        )
+
+    def draw_saved_bbox(
+        self,
+        bbox: List[float],
+        color: str,
+        width: int = 3,
+        label: Optional[str] = None,
+    ):
+        """
+        Draw a bounding box from database coordinates onto the annotation canvas.
+
+        Args:
+            bbox: Bounding box as [x_min_norm, y_min_norm, x_max_norm, y_max_norm]
+                  in normalized coordinates (0-1)
+            color: Color hex string (e.g., '#FF0000')
+            width: Line width in pixels
+            label: Optional text label to render near the bounding box
+        """
+        if not self.annotation_pixmap or not self.original_pixmap:
+            logger.warning("Cannot draw bounding box: no image loaded")
+            return
+
+        if len(bbox) != 4:
+            logger.warning(
+                f"Invalid bounding box format: expected 4 values, got {len(bbox)}"
+            )
+            return
+
+        # Convert normalized coordinates to image coordinates (for logging/debug)
+        img_width = self.original_pixmap.width()
+        img_height = self.original_pixmap.height()
+
+        x_min_norm, y_min_norm, x_max_norm, y_max_norm = bbox
+        x_min = int(x_min_norm * img_width)
+        y_min = int(y_min_norm * img_height)
+        x_max = int(x_max_norm * img_width)
+        y_max = int(y_max_norm * img_height)
+
+        logger.debug(f"Drawing bounding box: {bbox}")
+        logger.debug(f"  Image size: {img_width}x{img_height}")
+        logger.debug(f"  Pixel coords: ({x_min}, {y_min}) to ({x_max}, {y_max})")
+
+        # Store bounding box (normalized) and its style; actual drawing happens
+        # in _redraw_annotations() together with all polylines.
+        pen_color = QColor(color)
+        pen_color.setAlpha(128)  # Add semi-transparency
+        self.bboxes.append(
+            [float(x_min_norm), float(y_min_norm), float(x_max_norm), float(y_max_norm)]
+        )
+        self.bbox_meta.append({"color": pen_color, "width": int(width), "label": label})
+
+        # Store in all_strokes for consistency
+        self.all_strokes.append(
+            {"bbox": bbox, "color": color, "alpha": 128, "width": width, "label": label}
+        )
+
+        # Redraw overlay (polylines + all bounding boxes)
+        self._redraw_annotations()
+        logger.debug(f"Drew saved bounding box in color {color}")
+
+    def set_show_bboxes(self, show: bool):
+        """
+        Enable or disable drawing of bounding boxes.
+
+        Args:
+            show: If True, draw bounding boxes; if False, hide them.
+        """
+        self.show_bboxes = bool(show)
+        logger.debug(f"Set show_bboxes to {self.show_bboxes}")
+        self._redraw_annotations()
+
+    def keyPressEvent(self, event: QKeyEvent):
+        """Handle keyboard events for zooming."""
+        if event.key() in (Qt.Key_Plus, Qt.Key_Equal):
+            self.zoom_in()
+            event.accept()
+        elif event.key() == Qt.Key_Minus:
+            self.zoom_out()
+            event.accept()
+        elif event.key() == Qt.Key_0 and event.modifiers() == Qt.ControlModifier:
+            self.reset_zoom()
+            event.accept()
+        else:
+            super().keyPressEvent(event)
+
+    def eventFilter(self, obj, event: QEvent) -> bool:
+        """Event filter to capture wheel events for zooming."""
+        if event.type() == QEvent.Wheel:
+            wheel_event = event
+            if self.original_pixmap is not None:
+                delta = wheel_event.angleDelta().y()
+
+                if delta > 0:
+                    new_scale = self.zoom_scale + self.zoom_wheel_step
+                    if new_scale <= self.zoom_max:
+                        self.zoom_scale = new_scale
+                        self._apply_zoom()
+                else:
+                    new_scale = self.zoom_scale - self.zoom_wheel_step
+                    if new_scale >= self.zoom_min:
+                        self.zoom_scale = new_scale
+                        self._apply_zoom()
+
+                return True
+
+        return super().eventFilter(obj, event)

src/gui/widgets/annotation_tools_widget.py

@@ -0,0 +1,478 @@
+"""
+Annotation tools widget for controlling annotation parameters.
+Includes polyline tool, color picker, class selection, and annotation management.
+"""
+
+from PySide6.QtWidgets import (
+    QWidget,
+    QVBoxLayout,
+    QHBoxLayout,
+    QLabel,
+    QGroupBox,
+    QPushButton,
+    QComboBox,
+    QSpinBox,
+    QDoubleSpinBox,
+    QCheckBox,
+    QColorDialog,
+    QInputDialog,
+    QMessageBox,
+)
+from PySide6.QtGui import QColor, QIcon, QPixmap, QPainter
+from PySide6.QtCore import Qt, Signal
+from typing import Optional, Dict
+
+from src.database.db_manager import DatabaseManager
+from src.utils.logger import get_logger
+
+logger = get_logger(__name__)
+
+
+class AnnotationToolsWidget(QWidget):
+    """
+    Widget for annotation tool controls.
+
+    Features:
+    - Enable/disable polyline tool
+    - Color selection for polyline pen
+    - Object class selection
+    - Add new object classes
+    - Pen width control
+    - Clear annotations
+
+    Signals:
+        polyline_enabled_changed: Emitted when polyline tool is enabled/disabled (bool)
+        polyline_pen_color_changed: Emitted when polyline pen color changes (QColor)
+        polyline_pen_width_changed: Emitted when polyline pen width changes (int)
+        class_selected: Emitted when object class is selected (dict)
+        clear_annotations_requested: Emitted when clear button is pressed
+    """
+
+    polyline_enabled_changed = Signal(bool)
+    polyline_pen_color_changed = Signal(QColor)
+    polyline_pen_width_changed = Signal(int)
+    simplify_on_finish_changed = Signal(bool)
+    simplify_epsilon_changed = Signal(float)
+    # Toggle visibility of bounding boxes on the canvas
+    show_bboxes_changed = Signal(bool)
+    class_selected = Signal(dict)
+    class_color_changed = Signal()
+    clear_annotations_requested = Signal()
+    # Request deletion of the currently selected annotation on the canvas
+    delete_selected_annotation_requested = Signal()
+
+    def __init__(self, db_manager: DatabaseManager, parent=None):
+        """
+        Initialize annotation tools widget.
+
+        Args:
+            db_manager: Database manager instance
+            parent: Parent widget
+        """
+        super().__init__(parent)
+        self.db_manager = db_manager
+        self.polyline_enabled = False
+        self.current_color = QColor(255, 0, 0, 128)  # Red with 50% alpha
+        self.current_class = None
+
+        self._setup_ui()
+        self._load_object_classes()
+
+    def _setup_ui(self):
+        """Setup user interface."""
+        layout = QVBoxLayout()
+
+        # Polyline Tool Group
+        polyline_group = QGroupBox("Polyline Tool")
+        polyline_layout = QVBoxLayout()
+
+        # Enable/Disable polyline tool
+        button_layout = QHBoxLayout()
+        self.polyline_toggle_btn = QPushButton("Start Drawing Polyline")
+        self.polyline_toggle_btn.setCheckable(True)
+        self.polyline_toggle_btn.clicked.connect(self._on_polyline_toggle)
+        button_layout.addWidget(self.polyline_toggle_btn)
+        polyline_layout.addLayout(button_layout)
+
+        # Polyline pen width control
+        width_layout = QHBoxLayout()
+        width_layout.addWidget(QLabel("Pen Width:"))
+        self.polyline_pen_width_spin = QSpinBox()
+        self.polyline_pen_width_spin.setMinimum(1)
+        self.polyline_pen_width_spin.setMaximum(20)
+        self.polyline_pen_width_spin.setValue(3)
+        self.polyline_pen_width_spin.valueChanged.connect(
+            self._on_polyline_pen_width_changed
+        )
+        width_layout.addWidget(self.polyline_pen_width_spin)
+        width_layout.addStretch()
+        polyline_layout.addLayout(width_layout)
+
+        # Simplification controls (RDP)
+        simplify_layout = QHBoxLayout()
+        self.simplify_checkbox = QCheckBox("Simplify on finish")
+        self.simplify_checkbox.setChecked(True)
+        self.simplify_checkbox.stateChanged.connect(self._on_simplify_toggle)
+        simplify_layout.addWidget(self.simplify_checkbox)
+
+        simplify_layout.addWidget(QLabel("epsilon (px):"))
+        self.eps_spin = QDoubleSpinBox()
+        self.eps_spin.setRange(0.0, 1000.0)
+        self.eps_spin.setSingleStep(0.5)
+        self.eps_spin.setValue(2.0)
+        self.eps_spin.valueChanged.connect(self._on_eps_change)
+        simplify_layout.addWidget(self.eps_spin)
+        simplify_layout.addStretch()
+        polyline_layout.addLayout(simplify_layout)
+
+        polyline_group.setLayout(polyline_layout)
+        layout.addWidget(polyline_group)
+
+        # Object Class Group
+        class_group = QGroupBox("Object Class")
+        class_layout = QVBoxLayout()
+
+        # Class selection dropdown
+        self.class_combo = QComboBox()
+        self.class_combo.currentIndexChanged.connect(self._on_class_selected)
+        class_layout.addWidget(self.class_combo)
+
+        # Add / manage classes
+        class_button_layout = QHBoxLayout()
+        self.add_class_btn = QPushButton("Add New Class")
+        self.add_class_btn.clicked.connect(self._on_add_class)
+        class_button_layout.addWidget(self.add_class_btn)
+
+        self.refresh_classes_btn = QPushButton("Refresh")
+        self.refresh_classes_btn.clicked.connect(self._load_object_classes)
+        class_button_layout.addWidget(self.refresh_classes_btn)
+        class_layout.addLayout(class_button_layout)
+
+        # Class color (associated with selected object class)
+        color_layout = QHBoxLayout()
+        color_layout.addWidget(QLabel("Class Color:"))
+        self.color_btn = QPushButton()
+        self.color_btn.setFixedSize(40, 30)
+        self.color_btn.clicked.connect(self._on_color_picker)
+        self._update_color_button()
+        color_layout.addWidget(self.color_btn)
+        color_layout.addStretch()
+        class_layout.addLayout(color_layout)
+
+        # Selected class info
+        self.class_info_label = QLabel("No class selected")
+        self.class_info_label.setWordWrap(True)
+        self.class_info_label.setStyleSheet(
+            "QLabel { color: #888; font-style: italic; }"
+        )
+        class_layout.addWidget(self.class_info_label)
+
+        class_group.setLayout(class_layout)
+        layout.addWidget(class_group)
+
+        # Actions Group
+        actions_group = QGroupBox("Actions")
+        actions_layout = QVBoxLayout()
+
+        # Show / hide bounding boxes
+        self.show_bboxes_checkbox = QCheckBox("Show bounding boxes")
+        self.show_bboxes_checkbox.setChecked(True)
+        self.show_bboxes_checkbox.stateChanged.connect(self._on_show_bboxes_toggle)
+        actions_layout.addWidget(self.show_bboxes_checkbox)
+
+        self.clear_btn = QPushButton("Clear All Annotations")
+        self.clear_btn.clicked.connect(self._on_clear_annotations)
+        actions_layout.addWidget(self.clear_btn)
+
+        # Delete currently selected annotation (enabled when a selection exists)
+        self.delete_selected_btn = QPushButton("Delete Selected Annotation")
+        self.delete_selected_btn.clicked.connect(self._on_delete_selected_annotation)
+        self.delete_selected_btn.setEnabled(False)
+        actions_layout.addWidget(self.delete_selected_btn)
+
+        actions_group.setLayout(actions_layout)
+        layout.addWidget(actions_group)
+
+        layout.addStretch()
+        self.setLayout(layout)
+
+    def _update_color_button(self):
+        """Update the color button appearance with current color."""
+        pixmap = QPixmap(40, 30)
+        pixmap.fill(self.current_color)
+
+        # Add border
+        painter = QPainter(pixmap)
+        painter.setPen(Qt.black)
+        painter.drawRect(0, 0, pixmap.width() - 1, pixmap.height() - 1)
+        painter.end()
+
+        self.color_btn.setIcon(QIcon(pixmap))
+        self.color_btn.setStyleSheet(f"background-color: {self.current_color.name()};")
+
+    def _load_object_classes(self):
+        """Load object classes from database and populate combo box."""
+        try:
+            classes = self.db_manager.get_object_classes()
+
+            # Clear and repopulate combo box
+            self.class_combo.clear()
+            self.class_combo.addItem("-- Select Class / Show All --", None)
+
+            for cls in classes:
+                self.class_combo.addItem(cls["class_name"], cls)
+
+            logger.debug(f"Loaded {len(classes)} object classes")
+
+        except Exception as e:
+            logger.error(f"Error loading object classes: {e}")
+            QMessageBox.warning(
+                self, "Error", f"Failed to load object classes:\n{str(e)}"
+            )
+
+    def _on_polyline_toggle(self, checked: bool):
+        """Handle polyline tool enable/disable."""
+        self.polyline_enabled = checked
+
+        if checked:
+            self.polyline_toggle_btn.setText("Stop Drawing Polyline")
+            self.polyline_toggle_btn.setStyleSheet(
+                "QPushButton { background-color: #4CAF50; }"
+            )
+        else:
+            self.polyline_toggle_btn.setText("Start Drawing Polyline")
+            self.polyline_toggle_btn.setStyleSheet("")
+
+        self.polyline_enabled_changed.emit(self.polyline_enabled)
+        logger.debug(f"Polyline tool {'enabled' if checked else 'disabled'}")
+
+    def _on_polyline_pen_width_changed(self, width: int):
+        """Handle polyline pen width changes."""
+        self.polyline_pen_width_changed.emit(width)
+        logger.debug(f"Polyline pen width changed to {width}")
+
+    def _on_simplify_toggle(self, state: int):
+        """Handle simplify-on-finish checkbox toggle."""
+        enabled = bool(state)
+        self.simplify_on_finish_changed.emit(enabled)
+        logger.debug(f"Simplify on finish set to {enabled}")
+
+    def _on_eps_change(self, val: float):
+        """Handle epsilon (RDP tolerance) value changes."""
+        epsilon = float(val)
+        self.simplify_epsilon_changed.emit(epsilon)
+        logger.debug(f"Simplification epsilon changed to {epsilon}")
+
+    def _on_show_bboxes_toggle(self, state: int):
+        """Handle 'Show bounding boxes' checkbox toggle."""
+        show = bool(state)
+        self.show_bboxes_changed.emit(show)
+        logger.debug(f"Show bounding boxes set to {show}")
+
+    def _on_color_picker(self):
+        """Open color picker dialog and update the selected object's class color."""
+        if not self.current_class:
+            QMessageBox.warning(
+                self,
+                "No Class Selected",
+                "Please select an object class before changing its color.",
+            )
+            return
+
+        # Use current class color (without alpha) as the base
+        base_color = QColor(self.current_class.get("color", self.current_color.name()))
+        color = QColorDialog.getColor(
+            base_color,
+            self,
+            "Select Class Color",
+            QColorDialog.ShowAlphaChannel,  # Allow alpha in UI, but store RGB in DB
+        )
+
+        if not color.isValid():
+            return
+
+        # Normalize to opaque RGB for storage
+        new_color = QColor(color)
+        new_color.setAlpha(255)
+        hex_color = new_color.name()
+
+        try:
+            # Update in database
+            self.db_manager.update_object_class(
+                class_id=self.current_class["id"], color=hex_color
+            )
+        except Exception as e:
+            logger.error(f"Failed to update class color in database: {e}")
+            QMessageBox.critical(
+                self,
+                "Error",
+                f"Failed to update class color in database:\n{str(e)}",
+            )
+            return
+
+        # Update local class data and combo box item data
+        self.current_class["color"] = hex_color
+        current_index = self.class_combo.currentIndex()
+        if current_index >= 0:
+            self.class_combo.setItemData(current_index, dict(self.current_class))
+
+        # Update info label text
+        info_text = f"Class: {self.current_class['class_name']}\nColor: {hex_color}"
+        if self.current_class.get("description"):
+            info_text += f"\nDescription: {self.current_class['description']}"
+        self.class_info_label.setText(info_text)
+
+        # Use semi-transparent version for polyline pen / button preview
+        class_color = QColor(hex_color)
+        class_color.setAlpha(128)
+        self.current_color = class_color
+        self._update_color_button()
+        self.polyline_pen_color_changed.emit(class_color)
+
+        logger.debug(
+            f"Updated class '{self.current_class['class_name']}' color to "
+            f"{hex_color} (polyline pen alpha={class_color.alpha()})"
+        )
+
+        # Notify listeners (e.g., AnnotationTab) so they can reload/redraw
+        self.class_color_changed.emit()
+
+    def _on_class_selected(self, index: int):
+        """Handle object class selection (including '-- Select Class --')."""
+        class_data = self.class_combo.currentData()
+
+        if class_data:
+            self.current_class = class_data
+
+            # Update info label
+            info_text = (
+                f"Class: {class_data['class_name']}\n" f"Color: {class_data['color']}"
+            )
+            if class_data.get("description"):
+                info_text += f"\nDescription: {class_data['description']}"
+
+            self.class_info_label.setText(info_text)
+
+            # Update polyline pen color to match class color with semi-transparency
+            class_color = QColor(class_data["color"])
+            if class_color.isValid():
+                # Add 50% alpha for semi-transparency
+                class_color.setAlpha(128)
+                self.current_color = class_color
+                self._update_color_button()
+                self.polyline_pen_color_changed.emit(class_color)
+
+            self.class_selected.emit(class_data)
+            logger.debug(f"Selected class: {class_data['class_name']}")
+        else:
+            # "-- Select Class --" chosen: clear current class and show all annotations
+            self.current_class = None
+            self.class_info_label.setText("No class selected")
+            self.class_selected.emit(None)
+            logger.debug("Class selection cleared: showing annotations for all classes")
+
+    def _on_add_class(self):
+        """Handle adding a new object class."""
+        # Get class name
+        class_name, ok = QInputDialog.getText(
+            self, "Add Object Class", "Enter class name:"
+        )
+
+        if not ok or not class_name.strip():
+            return
+
+        class_name = class_name.strip()
+
+        # Check if class already exists
+        existing = self.db_manager.get_object_class_by_name(class_name)
+        if existing:
+            QMessageBox.warning(
+                self, "Class Exists", f"A class named '{class_name}' already exists."
+            )
+            return
+
+        # Get color
+        color = QColorDialog.getColor(self.current_color, self, "Select Class Color")
+
+        if not color.isValid():
+            return
+
+        # Get optional description
+        description, ok = QInputDialog.getText(
+            self, "Class Description", "Enter class description (optional):"
+        )
+
+        if not ok:
+            description = None
+
+        # Add to database
+        try:
+            class_id = self.db_manager.add_object_class(
+                class_name, color.name(), description.strip() if description else None
+            )
+
+            logger.info(f"Added new object class: {class_name} (ID: {class_id})")
+
+            # Reload classes and select the new one
+            self._load_object_classes()
+
+            # Find and select the newly added class
+            for i in range(self.class_combo.count()):
+                class_data = self.class_combo.itemData(i)
+                if class_data and class_data.get("id") == class_id:
+                    self.class_combo.setCurrentIndex(i)
+                    break
+
+            QMessageBox.information(
+                self, "Success", f"Class '{class_name}' added successfully!"
+            )
+
+        except Exception as e:
+            logger.error(f"Error adding object class: {e}")
+            QMessageBox.critical(
+                self, "Error", f"Failed to add object class:\n{str(e)}"
+            )
+
+    def _on_clear_annotations(self):
+        """Handle clear annotations button."""
+        reply = QMessageBox.question(
+            self,
+            "Clear Annotations",
+            "Are you sure you want to clear all annotations?",
+            QMessageBox.Yes | QMessageBox.No,
+            QMessageBox.No,
+        )
+
+        if reply == QMessageBox.Yes:
+            self.clear_annotations_requested.emit()
+            logger.debug("Clear annotations requested")
+
+    def _on_delete_selected_annotation(self):
+        """Handle delete selected annotation button."""
+        self.delete_selected_annotation_requested.emit()
+        logger.debug("Delete selected annotation requested")
+
+    def set_has_selected_annotation(self, has_selection: bool):
+        """
+        Enable/disable actions that require a selected annotation.
+
+        Args:
+            has_selection: True if an annotation is currently selected on the canvas.
+        """
+        self.delete_selected_btn.setEnabled(bool(has_selection))
+
+    def get_current_class(self) -> Optional[Dict]:
+        """Get currently selected object class."""
+        return self.current_class
+
+    def get_polyline_pen_color(self) -> QColor:
+        """Get current polyline pen color."""
+        return self.current_color
+
+    def get_polyline_pen_width(self) -> int:
+        """Get current polyline pen width."""
+        return self.polyline_pen_width_spin.value()
+
+    def is_polyline_enabled(self) -> bool:
+        """Check if polyline tool is enabled."""
+        return self.polyline_enabled

src/gui/widgets/image_display_widget.py

@@ -0,0 +1,282 @@
+"""
+Image display widget with zoom functionality for the microscopy object detection application.
+Reusable widget for displaying images with zoom controls.
+"""
+
+from PySide6.QtWidgets import QWidget, QVBoxLayout, QLabel, QScrollArea
+from PySide6.QtGui import QPixmap, QImage, QKeyEvent
+from PySide6.QtCore import Qt, QEvent, Signal
+from pathlib import Path
+import numpy as np
+
+from src.utils.image import Image, ImageLoadError
+from src.utils.logger import get_logger
+
+logger = get_logger(__name__)
+
+
+class ImageDisplayWidget(QWidget):
+    """
+    Reusable widget for displaying images with zoom functionality.
+
+    Features:
+    - Display images from Image objects
+    - Zoom in/out with mouse wheel
+    - Zoom in/out with +/- keyboard keys
+    - Reset zoom with Ctrl+0
+    - Scroll area for large images
+
+    Signals:
+        zoom_changed: Emitted when zoom level changes (float zoom_scale)
+    """
+
+    zoom_changed = Signal(float)  # Emitted when zoom level changes
+
+    def __init__(self, parent=None):
+        """
+        Initialize the image display widget.
+
+        Args:
+            parent: Parent widget
+        """
+        super().__init__(parent)
+
+        self.current_image = None
+        self.original_pixmap = None  # Store original pixmap for zoom
+        self.zoom_scale = 1.0  # Current zoom scale
+        self.zoom_min = 0.1  # Minimum zoom (10%)
+        self.zoom_max = 10.0  # Maximum zoom (1000%)
+        self.zoom_step = 0.1  # Zoom step for +/- keys
+        self.zoom_wheel_step = 0.15  # Zoom step for mouse wheel
+
+        self._setup_ui()
+
+    def _setup_ui(self):
+        """Setup user interface."""
+        layout = QVBoxLayout()
+        layout.setContentsMargins(0, 0, 0, 0)
+
+        # Scroll area for image
+        self.scroll_area = QScrollArea()
+        self.scroll_area.setWidgetResizable(True)
+        self.scroll_area.setMinimumHeight(400)
+
+        self.image_label = QLabel("No image loaded")
+        self.image_label.setAlignment(Qt.AlignCenter)
+        self.image_label.setStyleSheet(
+            "QLabel { background-color: #2b2b2b; color: #888; }"
+        )
+        self.image_label.setScaledContents(False)
+
+        # Enable mouse tracking for wheel events
+        self.image_label.setMouseTracking(True)
+        self.scroll_area.setWidget(self.image_label)
+
+        # Install event filter to capture wheel events on scroll area
+        self.scroll_area.viewport().installEventFilter(self)
+
+        layout.addWidget(self.scroll_area)
+        self.setLayout(layout)
+
+        # Set focus policy to receive keyboard events
+        self.setFocusPolicy(Qt.StrongFocus)
+
+    def load_image(self, image: Image):
+        """
+        Load and display an image.
+
+        Args:
+            image: Image object to display
+
+        Raises:
+            ImageLoadError: If image cannot be displayed
+        """
+        self.current_image = image
+
+        # Reset zoom when loading new image
+        self.zoom_scale = 1.0
+
+        # Convert to QPixmap and display
+        self._display_image()
+
+        logger.debug(f"Loaded image into display widget: {image.width}x{image.height}")
+
+    def clear(self):
+        """Clear the displayed image."""
+        self.current_image = None
+        self.original_pixmap = None
+        self.zoom_scale = 1.0
+        self.image_label.setText("No image loaded")
+        self.image_label.setPixmap(QPixmap())
+        logger.debug("Cleared image display")
+
+    def _display_image(self):
+        """Display the current image in the image label."""
+        if self.current_image is None:
+            return
+
+        try:
+            # Get RGB image data
+            if self.current_image.channels == 3:
+                image_data = self.current_image.get_rgb()
+                height, width, channels = image_data.shape
+            else:
+                image_data = self.current_image.get_grayscale()
+                height, width = image_data.shape
+                channels = 1
+
+            # Ensure data is contiguous for proper QImage display
+            image_data = np.ascontiguousarray(image_data)
+
+            # Use actual stride from numpy array for correct display
+            bytes_per_line = image_data.strides[0]
+
+            qimage = QImage(
+                image_data.data,
+                width,
+                height,
+                bytes_per_line,
+                self.current_image.qtimage_format,
+            ).copy()  # Copy to ensure Qt owns its memory after this scope
+
+            # Convert to pixmap
+            pixmap = QPixmap.fromImage(qimage)
+
+            # Store original pixmap for zooming
+            self.original_pixmap = pixmap
+
+            # Apply zoom and display
+            self._apply_zoom()
+
+        except Exception as e:
+            logger.error(f"Error displaying image: {e}")
+            raise ImageLoadError(f"Failed to display image: {str(e)}")
+
+    def _apply_zoom(self):
+        """Apply current zoom level to the displayed image."""
+        if self.original_pixmap is None:
+            return
+
+        # Calculate scaled size
+        scaled_width = int(self.original_pixmap.width() * self.zoom_scale)
+        scaled_height = int(self.original_pixmap.height() * self.zoom_scale)
+
+        # Scale pixmap
+        scaled_pixmap = self.original_pixmap.scaled(
+            scaled_width,
+            scaled_height,
+            Qt.KeepAspectRatio,
+            (
+                Qt.SmoothTransformation
+                if self.zoom_scale >= 1.0
+                else Qt.FastTransformation
+            ),
+        )
+
+        # Display in label
+        self.image_label.setPixmap(scaled_pixmap)
+        self.image_label.setScaledContents(False)
+        self.image_label.adjustSize()
+
+        # Emit zoom changed signal
+        self.zoom_changed.emit(self.zoom_scale)
+
+    def zoom_in(self):
+        """Zoom in on the image."""
+        if self.original_pixmap is None:
+            return
+
+        new_scale = self.zoom_scale + self.zoom_step
+        if new_scale <= self.zoom_max:
+            self.zoom_scale = new_scale
+            self._apply_zoom()
+            logger.debug(f"Zoomed in to {int(self.zoom_scale * 100)}%")
+
+    def zoom_out(self):
+        """Zoom out from the image."""
+        if self.original_pixmap is None:
+            return
+
+        new_scale = self.zoom_scale - self.zoom_step
+        if new_scale >= self.zoom_min:
+            self.zoom_scale = new_scale
+            self._apply_zoom()
+            logger.debug(f"Zoomed out to {int(self.zoom_scale * 100)}%")
+
+    def reset_zoom(self):
+        """Reset zoom to 100%."""
+        if self.original_pixmap is None:
+            return
+
+        self.zoom_scale = 1.0
+        self._apply_zoom()
+        logger.debug("Reset zoom to 100%")
+
+    def set_zoom(self, scale: float):
+        """
+        Set zoom to a specific scale.
+
+        Args:
+            scale: Zoom scale (1.0 = 100%)
+        """
+        if self.original_pixmap is None:
+            return
+
+        # Clamp to min/max
+        scale = max(self.zoom_min, min(self.zoom_max, scale))
+
+        self.zoom_scale = scale
+        self._apply_zoom()
+        logger.debug(f"Set zoom to {int(self.zoom_scale * 100)}%")
+
+    def get_zoom_percentage(self) -> int:
+        """
+        Get current zoom level as percentage.
+
+        Returns:
+            Zoom level as integer percentage (e.g., 100 for 100%)
+        """
+        return int(self.zoom_scale * 100)
+
+    def keyPressEvent(self, event: QKeyEvent):
+        """Handle keyboard events for zooming."""
+        if event.key() in (Qt.Key_Plus, Qt.Key_Equal):
+            # + or = key (= is the unshifted + on many keyboards)
+            self.zoom_in()
+            event.accept()
+        elif event.key() == Qt.Key_Minus:
+            # - key
+            self.zoom_out()
+            event.accept()
+        elif event.key() == Qt.Key_0 and event.modifiers() == Qt.ControlModifier:
+            # Ctrl+0 to reset zoom
+            self.reset_zoom()
+            event.accept()
+        else:
+            super().keyPressEvent(event)
+
+    def eventFilter(self, obj, event: QEvent) -> bool:
+        """Event filter to capture wheel events for zooming."""
+        if event.type() == QEvent.Wheel:
+            wheel_event = event
+            if self.original_pixmap is not None:
+                # Get wheel angle delta
+                delta = wheel_event.angleDelta().y()
+
+                # Zoom in/out based on wheel direction
+                if delta > 0:
+                    # Scroll up = zoom in
+                    new_scale = self.zoom_scale + self.zoom_wheel_step
+                    if new_scale <= self.zoom_max:
+                        self.zoom_scale = new_scale
+                        self._apply_zoom()
+                else:
+                    # Scroll down = zoom out
+                    new_scale = self.zoom_scale - self.zoom_wheel_step
+                    if new_scale >= self.zoom_min:
+                        self.zoom_scale = new_scale
+                        self._apply_zoom()
+
+                return True  # Event handled
+
+        return super().eventFilter(obj, event)

src/gui_launcher.py

@@ -0,0 +1,49 @@
+"""GUI launcher module for microscopy object detection application."""
+
+import sys
+from pathlib import Path
+
+from PySide6.QtWidgets import QApplication
+from PySide6.QtCore import Qt
+
+from src import __version__
+from src.gui.main_window import MainWindow
+from src.utils.logger import setup_logging
+from src.utils.config_manager import ConfigManager
+
+
+def main():
+    """Launch the GUI application."""
+    # Setup logging
+    config_manager = ConfigManager()
+    log_config = config_manager.get_section("logging")
+    setup_logging(
+        log_file=log_config.get("file", "logs/app.log"),
+        level=log_config.get("level", "INFO"),
+        log_format=log_config.get("format"),
+    )
+
+    # Enable High DPI scaling
+    QApplication.setHighDpiScaleFactorRoundingPolicy(
+        Qt.HighDpiScaleFactorRoundingPolicy.PassThrough
+    )
+
+    # Create Qt application
+    app = QApplication(sys.argv)
+    app.setApplicationName("Microscopy Object Detection")
+    app.setOrganizationName("MicroscopyLab")
+    app.setApplicationVersion(__version__)
+
+    # Set application style
+    app.setStyle("Fusion")
+
+    # Create and show main window
+    window = MainWindow()
+    window.show()
+
+    # Run application
+    sys.exit(app.exec())
+
+
+if __name__ == "__main__":
+    main()

src/model/inference.py

@@ -0,0 +1,384 @@
+"""
+Inference engine for the microscopy object detection application.
+Handles detection inference and result storage.
+"""
+
+from typing import List, Dict, Optional, Callable
+from pathlib import Path
+import cv2
+import numpy as np
+
+from src.model.yolo_wrapper import YOLOWrapper
+from src.database.db_manager import DatabaseManager
+from src.utils.image import Image
+from src.utils.logger import get_logger
+from src.utils.file_utils import get_relative_path
+
+
+logger = get_logger(__name__)
+
+
+class InferenceEngine:
+    """Handles detection inference and result storage."""
+
+    def __init__(self, model_path: str, db_manager: DatabaseManager, model_id: int):
+        """
+        Initialize inference engine.
+
+        Args:
+            model_path: Path to YOLO model weights
+            db_manager: Database manager instance
+            model_id: ID of the model in database
+        """
+        self.yolo = YOLOWrapper(model_path)
+        self.yolo.load_model()
+        self.db_manager = db_manager
+        self.model_id = model_id
+        logger.info(f"InferenceEngine initialized with model_id {model_id}")
+
+    def detect_single(
+        self,
+        image_path: str,
+        relative_path: str,
+        conf: float = 0.25,
+        save_to_db: bool = True,
+        repository_root: Optional[str] = None,
+    ) -> Dict:
+        """
+        Detect objects in a single image.
+
+        Args:
+            image_path: Absolute path to image file
+            relative_path: Relative path from repository root
+            conf: Confidence threshold
+            save_to_db: Whether to save results to database
+            repository_root: Base directory used to compute relative_path (if known)
+
+        Returns:
+            Dictionary with detection results
+        """
+        try:
+            # Normalize storage path (fall back to absolute path when repo root is unknown)
+            stored_relative_path = relative_path
+            if not repository_root:
+                stored_relative_path = str(Path(image_path).resolve())
+
+            # Get image dimensions
+            img = Image(image_path)
+            width = img.width
+            height = img.height
+
+            # Perform detection
+            detections = self.yolo.predict(image_path, conf=conf)
+
+            # Add/get image in database
+            image_id = self.db_manager.get_or_create_image(
+                relative_path=stored_relative_path,
+                filename=Path(image_path).name,
+                width=width,
+                height=height,
+            )
+
+            inserted_count = 0
+            deleted_count = 0
+
+            # Save detections to database, replacing any previous results for this image/model
+            if save_to_db:
+                deleted_count = self.db_manager.delete_detections_for_image(
+                    image_id, self.model_id
+                )
+                if detections:
+                    detection_records = []
+                    for det in detections:
+                        # Use normalized bbox from detection
+                        bbox_normalized = det[
+                            "bbox_normalized"
+                        ]  # [x_min, y_min, x_max, y_max]
+
+                        metadata = {
+                            "class_id": det["class_id"],
+                            "source_path": str(Path(image_path).resolve()),
+                        }
+                        if repository_root:
+                            metadata["repository_root"] = str(
+                                Path(repository_root).resolve()
+                            )
+
+                        record = {
+                            "image_id": image_id,
+                            "model_id": self.model_id,
+                            "class_name": det["class_name"],
+                            "bbox": tuple(bbox_normalized),
+                            "confidence": det["confidence"],
+                            "segmentation_mask": det.get("segmentation_mask"),
+                            "metadata": metadata,
+                        }
+                        detection_records.append(record)
+
+                    inserted_count = self.db_manager.add_detections_batch(
+                        detection_records
+                    )
+                    logger.info(
+                        f"Saved {inserted_count} detections to database (replaced {deleted_count})"
+                    )
+                else:
+                    logger.info(
+                        f"Detection run removed {deleted_count} stale entries but produced no new detections"
+                    )
+
+            return {
+                "success": True,
+                "image_path": image_path,
+                "image_id": image_id,
+                "detections": detections,
+                "count": len(detections),
+            }
+
+        except Exception as e:
+            logger.error(f"Error detecting objects in {image_path}: {e}")
+            return {
+                "success": False,
+                "image_path": image_path,
+                "error": str(e),
+                "detections": [],
+                "count": 0,
+            }
+
+    def detect_batch(
+        self,
+        image_paths: List[str],
+        repository_root: str,
+        conf: float = 0.25,
+        progress_callback: Optional[Callable[[int, int, str], None]] = None,
+    ) -> List[Dict]:
+        """
+        Detect objects in multiple images.
+
+        Args:
+            image_paths: List of absolute image paths
+            repository_root: Root directory for relative paths
+            conf: Confidence threshold
+            progress_callback: Optional callback(current, total, message)
+
+        Returns:
+            List of detection result dictionaries
+        """
+        results = []
+        total = len(image_paths)
+
+        logger.info(f"Starting batch detection on {total} images")
+
+        for i, image_path in enumerate(image_paths, 1):
+            # Calculate relative path
+            rel_path = get_relative_path(image_path, repository_root)
+
+            # Perform detection
+            result = self.detect_single(
+                image_path,
+                rel_path,
+                conf=conf,
+                repository_root=repository_root,
+            )
+            results.append(result)
+
+            # Update progress
+            if progress_callback:
+                progress_callback(i, total, f"Processed {rel_path}")
+
+            if i % 10 == 0:
+                logger.info(f"Processed {i}/{total} images")
+
+        logger.info(f"Batch detection complete: {total} images processed")
+        return results
+
+    def detect_with_visualization(
+        self,
+        image_path: str,
+        conf: float = 0.25,
+        bbox_thickness: int = 2,
+        bbox_colors: Optional[Dict[str, str]] = None,
+        draw_masks: bool = True,
+    ) -> tuple:
+        """
+        Detect objects and return annotated image.
+
+        Args:
+            image_path: Path to image
+            conf: Confidence threshold
+            bbox_thickness: Thickness of bounding boxes
+            bbox_colors: Dictionary mapping class names to hex colors
+            draw_masks: Whether to draw segmentation masks (if available)
+
+        Returns:
+            Tuple of (detections, annotated_image_array)
+        """
+        try:
+            detections = self.yolo.predict(image_path, conf=conf)
+
+            # Load image
+            img = cv2.imread(image_path)
+            if img is None:
+                raise ValueError(f"Failed to load image: {image_path}")
+
+            img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
+            height, width = img.shape[:2]
+
+            # Default colors if not provided
+            if bbox_colors is None:
+                bbox_colors = {}
+            default_color = self._hex_to_bgr(bbox_colors.get("default", "#00FF00"))
+
+            # Draw detections
+            for det in detections:
+                # Get color for this class
+                class_name = det["class_name"]
+                color_hex = bbox_colors.get(
+                    class_name, bbox_colors.get("default", "#00FF00")
+                )
+                color = self._hex_to_bgr(color_hex)
+
+                # Draw segmentation mask if available and requested
+                if draw_masks and det.get("segmentation_mask"):
+                    mask_normalized = det["segmentation_mask"]
+                    if mask_normalized and len(mask_normalized) > 0:
+                        # Convert normalized coordinates to absolute pixels
+                        mask_points = np.array(
+                            [
+                                [int(pt[0] * width), int(pt[1] * height)]
+                                for pt in mask_normalized
+                            ],
+                            dtype=np.int32,
+                        )
+
+                        # Create a semi-transparent overlay
+                        overlay = img.copy()
+                        cv2.fillPoly(overlay, [mask_points], color)
+                        # Blend with original image (30% opacity)
+                        cv2.addWeighted(overlay, 0.3, img, 0.7, 0, img)
+
+                        # Draw mask contour
+                        cv2.polylines(img, [mask_points], True, color, bbox_thickness)
+
+                # Get absolute coordinates for bounding box
+                bbox_abs = det["bbox_absolute"]
+                x1, y1, x2, y2 = [int(v) for v in bbox_abs]
+
+                # Draw bounding box
+                cv2.rectangle(img, (x1, y1), (x2, y2), color, bbox_thickness)
+
+                # Prepare label
+                label = f"{class_name} {det['confidence']:.2f}"
+
+                # Draw label background
+                (label_w, label_h), baseline = cv2.getTextSize(
+                    label, cv2.FONT_HERSHEY_SIMPLEX, 0.5, 1
+                )
+                cv2.rectangle(
+                    img,
+                    (x1, y1 - label_h - baseline - 5),
+                    (x1 + label_w, y1),
+                    color,
+                    -1,
+                )
+
+                # Draw label text
+                cv2.putText(
+                    img,
+                    label,
+                    (x1, y1 - baseline - 5),
+                    cv2.FONT_HERSHEY_SIMPLEX,
+                    0.5,
+                    (255, 255, 255),
+                    1,
+                )
+
+            return detections, img
+
+        except Exception as e:
+            logger.error(f"Error creating visualization: {e}")
+            # Return empty detections and original image if possible
+            try:
+                img = cv2.imread(image_path)
+                img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
+                return [], img
+            except:
+                return [], np.zeros((480, 640, 3), dtype=np.uint8)
+
+    def get_detection_summary(self, detections: List[Dict]) -> Dict[str, any]:
+        """
+        Generate summary statistics for detections.
+
+        Args:
+            detections: List of detection dictionaries
+
+        Returns:
+            Dictionary with summary statistics
+        """
+        if not detections:
+            return {
+                "total_count": 0,
+                "class_counts": {},
+                "avg_confidence": 0.0,
+                "confidence_range": (0.0, 0.0),
+            }
+
+        # Count by class
+        class_counts = {}
+        confidences = []
+
+        for det in detections:
+            class_name = det["class_name"]
+            class_counts[class_name] = class_counts.get(class_name, 0) + 1
+            confidences.append(det["confidence"])
+
+        return {
+            "total_count": len(detections),
+            "class_counts": class_counts,
+            "avg_confidence": sum(confidences) / len(confidences),
+            "confidence_range": (min(confidences), max(confidences)),
+        }
+
+    @staticmethod
+    def _hex_to_bgr(hex_color: str) -> tuple:
+        """
+        Convert hex color to BGR tuple.
+
+        Args:
+            hex_color: Hex color string (e.g., '#FF0000')
+
+        Returns:
+            BGR tuple (B, G, R)
+        """
+        hex_color = hex_color.lstrip("#")
+        if len(hex_color) != 6:
+            return (0, 255, 0)  # Default green
+
+        try:
+            r = int(hex_color[0:2], 16)
+            g = int(hex_color[2:4], 16)
+            b = int(hex_color[4:6], 16)
+            return (b, g, r)  # OpenCV uses BGR
+        except ValueError:
+            return (0, 255, 0)  # Default green
+
+    def change_model(self, model_path: str, model_id: int) -> bool:
+        """
+        Change the current model.
+
+        Args:
+            model_path: Path to new model weights
+            model_id: ID of new model in database
+
+        Returns:
+            True if successful, False otherwise
+        """
+        try:
+            self.yolo = YOLOWrapper(model_path)
+            if self.yolo.load_model():
+                self.model_id = model_id
+                logger.info(f"Model changed to {model_path}")
+                return True
+            return False
+        except Exception as e:
+            logger.error(f"Error changing model: {e}")
+            return False

src/model/yolo_wrapper.py

@@ -0,0 +1,519 @@
+"""
+YOLO model wrapper for the microscopy object detection application.
+Provides a clean interface to YOLOv8 for training, validation, and inference.
+"""
+
+from ultralytics import YOLO
+from pathlib import Path
+from typing import Optional, List, Dict, Callable, Any
+import torch
+import tempfile
+import os
+import numpy as np
+from src.utils.image import Image, convert_grayscale_to_rgb_preserve_range
+from src.utils.logger import get_logger
+from src.utils.train_ultralytics_float import train_with_float32_loader
+
+
+logger = get_logger(__name__)
+
+
+class YOLOWrapper:
+    """Wrapper for YOLOv8 model operations."""
+
+    def __init__(self, model_path: str = "yolov8s-seg.pt"):
+        """
+        Initialize YOLO model.
+
+        Args:
+            model_path: Path to model weights (.pt file)
+        """
+        self.model_path = model_path
+        self.model = None
+        self.device = "cuda" if torch.cuda.is_available() else "cpu"
+        logger.info(f"YOLOWrapper initialized with device: {self.device}")
+
+    def load_model(self) -> bool:
+        """
+        Load YOLO model from path.
+
+        Returns:
+            True if loaded successfully, False otherwise
+        """
+        try:
+            logger.info(f"Loading YOLO model from {self.model_path}")
+            self.model = YOLO(self.model_path)
+            self.model.to(self.device)
+            logger.info("Model loaded successfully")
+            return True
+        except Exception as e:
+            logger.error(f"Error loading model: {e}")
+            return False
+
+    def train(
+        self,
+        data_yaml: str,
+        epochs: int = 100,
+        imgsz: int = 640,
+        batch: int = 16,
+        patience: int = 50,
+        save_dir: str = "data/models",
+        name: str = "custom_model",
+        resume: bool = False,
+        callbacks: Optional[Dict[str, Callable]] = None,
+        use_float32_loader: bool = True,
+        **kwargs,
+    ) -> Dict[str, Any]:
+        """
+        Train the YOLO model with optional float32 loader for 16-bit TIFFs.
+
+        Args:
+            data_yaml: Path to data.yaml configuration file
+            epochs: Number of training epochs
+            imgsz: Input image size
+            batch: Batch size
+            patience: Early stopping patience
+            save_dir: Directory to save trained model
+            name: Name for the training run
+            resume: Resume training from last checkpoint
+            callbacks: Optional Ultralytics callback dictionary
+            use_float32_loader: Use custom Float32Dataset for 16-bit TIFFs (default: True)
+            **kwargs: Additional training arguments
+
+        Returns:
+            Dictionary with training results
+        """
+        if 1:
+            logger.info(f"Starting training: {name}")
+            logger.info(
+                f"Data: {data_yaml}, Epochs: {epochs}, Batch: {batch}, ImgSz: {imgsz}"
+            )
+
+            # Check if dataset has 16-bit TIFFs and use float32 loader
+            if use_float32_loader:
+                logger.info("Using Float32Dataset loader for 16-bit TIFF support")
+                return train_with_float32_loader(
+                    model_path=self.model_path,
+                    data_yaml=data_yaml,
+                    epochs=epochs,
+                    imgsz=imgsz,
+                    batch=batch,
+                    patience=patience,
+                    save_dir=save_dir,
+                    name=name,
+                    callbacks=callbacks,
+                    device=self.device,
+                    resume=resume,
+                    **kwargs,
+                )
+            else:
+                # Standard training (old behavior)
+                if self.model is None:
+                    if not self.load_model():
+                        raise RuntimeError(
+                            f"Failed to load model from {self.model_path}"
+                        )
+
+                results = self.model.train(
+                    data=data_yaml,
+                    epochs=epochs,
+                    imgsz=imgsz,
+                    batch=batch,
+                    patience=patience,
+                    project=save_dir,
+                    name=name,
+                    device=self.device,
+                    resume=resume,
+                    **kwargs,
+                )
+
+                logger.info("Training completed successfully")
+                return self._format_training_results(results)
+
+        # except Exception as e:
+        #     logger.error(f"Error during training: {e}")
+        #     raise
+
+    def validate(self, data_yaml: str, split: str = "val", **kwargs) -> Dict[str, Any]:
+        """
+        Validate the model.
+
+        Args:
+            data_yaml: Path to data.yaml configuration file
+            split: Dataset split to validate on ('val' or 'test')
+            **kwargs: Additional validation arguments
+
+        Returns:
+            Dictionary with validation metrics
+        """
+        if self.model is None:
+            if not self.load_model():
+                raise RuntimeError(f"Failed to load model from {self.model_path}")
+
+        try:
+            logger.info(f"Starting validation on {split} split")
+            results = self.model.val(
+                data=data_yaml, split=split, device=self.device, **kwargs
+            )
+
+            logger.info("Validation completed successfully")
+            return self._format_validation_results(results)
+
+        except Exception as e:
+            logger.error(f"Error during validation: {e}")
+            raise
+
+    def predict(
+        self,
+        source: str,
+        conf: float = 0.25,
+        iou: float = 0.45,
+        save: bool = False,
+        save_txt: bool = False,
+        save_conf: bool = False,
+        **kwargs,
+    ) -> List[Dict]:
+        """
+        Perform inference on image(s).
+
+        Args:
+            source: Path to image or directory
+            conf: Confidence threshold
+            iou: IoU threshold for NMS
+            save: Whether to save annotated images
+            save_txt: Whether to save labels to .txt files
+            save_conf: Whether to save confidence in labels
+            **kwargs: Additional prediction arguments
+
+        Returns:
+            List of detection dictionaries
+        """
+        if self.model is None:
+            if not self.load_model():
+                raise RuntimeError(f"Failed to load model from {self.model_path}")
+
+        prepared_source, cleanup_path = self._prepare_source(source)
+
+        try:
+            logger.info(f"Running inference on {source}")
+            results = self.model.predict(
+                source=prepared_source,
+                conf=conf,
+                iou=iou,
+                save=save,
+                save_txt=save_txt,
+                save_conf=save_conf,
+                device=self.device,
+                **kwargs,
+            )
+
+            detections = self._format_prediction_results(results)
+            logger.info(f"Inference complete: {len(detections)} detections")
+            return detections
+
+        except Exception as e:
+            logger.error(f"Error during inference: {e}")
+            raise
+        finally:
+            # Clean up temporary files (only for non-16-bit images)
+            # 16-bit TIFFs return numpy arrays directly, so cleanup_path is None
+            if cleanup_path:
+                try:
+                    os.remove(cleanup_path)
+                    logger.debug(f"Cleaned up temporary file: {cleanup_path}")
+                except OSError as cleanup_error:
+                    logger.warning(
+                        f"Failed to delete temporary file {cleanup_path}: {cleanup_error}"
+                    )
+
+    def export(
+        self, format: str = "onnx", output_path: Optional[str] = None, **kwargs
+    ) -> str:
+        """
+        Export model to different format.
+
+        Args:
+            format: Export format (onnx, torchscript, tflite, etc.)
+            output_path: Path for exported model
+            **kwargs: Additional export arguments
+
+        Returns:
+            Path to exported model
+        """
+        if self.model is None:
+            if not self.load_model():
+                raise RuntimeError(f"Failed to load model from {self.model_path}")
+
+        try:
+            logger.info(f"Exporting model to {format} format")
+            export_path = self.model.export(format=format, **kwargs)
+            logger.info(f"Model exported to {export_path}")
+            return str(export_path)
+
+        except Exception as e:
+            logger.error(f"Error exporting model: {e}")
+            raise
+
+    def _prepare_source(self, source):
+        """Convert single-channel images to RGB for inference.
+
+        For 16-bit TIFF files, this will:
+        1. Load using tifffile
+        2. Normalize to float32 [0-1] (NO uint8 conversion to avoid data loss)
+        3. Replicate grayscale → RGB (3 channels)
+        4. Pass directly as numpy array to YOLO
+        """
+        cleanup_path = None
+
+        if isinstance(source, (str, Path)):
+            source_path = Path(source)
+            if source_path.is_file():
+                try:
+                    img_obj = Image(source_path)
+
+                    # Check if it's a 16-bit TIFF file
+                    is_16bit_tiff = (
+                        source_path.suffix.lower() in [".tif", ".tiff"]
+                        and img_obj.dtype == np.uint16
+                    )
+
+                    if is_16bit_tiff:
+                        # Process 16-bit TIFF: normalize to float32 [0-1]
+                        # NO uint8 conversion - pass float32 directly to avoid data loss
+                        normalized_float = img_obj.to_normalized_float32()
+
+                        # Convert grayscale to RGB by replicating channels
+                        if len(normalized_float.shape) == 2:
+                            # Grayscale: H,W → H,W,3
+                            rgb_float = np.stack([normalized_float] * 3, axis=-1)
+                        elif (
+                            len(normalized_float.shape) == 3
+                            and normalized_float.shape[2] == 1
+                        ):
+                            # Grayscale with channel dim: H,W,1 → H,W,3
+                            rgb_float = np.repeat(normalized_float, 3, axis=2)
+                        else:
+                            # Already multi-channel
+                            rgb_float = normalized_float
+
+                        # Ensure contiguous array and float32
+                        rgb_float = np.ascontiguousarray(rgb_float, dtype=np.float32)
+
+                        logger.info(
+                            f"Loaded 16-bit TIFF {source_path} as float32 [0-1] RGB "
+                            f"(shape: {rgb_float.shape}, dtype: {rgb_float.dtype}, "
+                            f"range: [{rgb_float.min():.4f}, {rgb_float.max():.4f}])"
+                        )
+
+                        # Return numpy array directly - YOLO can handle it
+                        return rgb_float, cleanup_path
+                    else:
+                        # Standard processing for other images
+                        pil_img = img_obj.pil_image
+                        if len(pil_img.getbands()) == 1:
+                            rgb_img = convert_grayscale_to_rgb_preserve_range(pil_img)
+                        else:
+                            rgb_img = pil_img.convert("RGB")
+
+                        suffix = source_path.suffix or ".png"
+                        tmp = tempfile.NamedTemporaryFile(suffix=suffix, delete=False)
+                        tmp_path = tmp.name
+                        tmp.close()
+                        rgb_img.save(tmp_path)
+                        cleanup_path = tmp_path
+                        logger.info(
+                            f"Converted image {source_path} to RGB for inference at {tmp_path}"
+                        )
+                        return tmp_path, cleanup_path
+                except Exception as convert_error:
+                    logger.warning(
+                        f"Failed to preprocess {source_path} as RGB, continuing with original file: {convert_error}"
+                    )
+
+        return source, cleanup_path
+
+    def _format_training_results(self, results) -> Dict[str, Any]:
+        """Format training results into dictionary."""
+        try:
+            # Get the results dict
+            results_dict = (
+                results.results_dict if hasattr(results, "results_dict") else {}
+            )
+
+            formatted = {
+                "success": True,
+                "final_epoch": getattr(results, "epoch", 0),
+                "metrics": {
+                    "mAP50": float(results_dict.get("metrics/mAP50(B)", 0)),
+                    "mAP50-95": float(results_dict.get("metrics/mAP50-95(B)", 0)),
+                    "precision": float(results_dict.get("metrics/precision(B)", 0)),
+                    "recall": float(results_dict.get("metrics/recall(B)", 0)),
+                },
+                "best_model_path": str(Path(results.save_dir) / "weights" / "best.pt"),
+                "last_model_path": str(Path(results.save_dir) / "weights" / "last.pt"),
+                "save_dir": str(results.save_dir),
+            }
+
+            return formatted
+
+        except Exception as e:
+            logger.error(f"Error formatting training results: {e}")
+            return {"success": False, "error": str(e)}
+
+    def _format_validation_results(self, results) -> Dict[str, Any]:
+        """Format validation results into dictionary."""
+        try:
+            box_metrics = results.box
+
+            formatted = {
+                "success": True,
+                "mAP50": float(box_metrics.map50),
+                "mAP50-95": float(box_metrics.map),
+                "precision": float(box_metrics.mp),
+                "recall": float(box_metrics.mr),
+                "fitness": (
+                    float(results.fitness) if hasattr(results, "fitness") else 0.0
+                ),
+            }
+
+            # Add per-class metrics if available
+            if hasattr(box_metrics, "ap") and hasattr(results, "names"):
+                class_metrics = {}
+                for idx, name in results.names.items():
+                    if idx < len(box_metrics.ap):
+                        class_metrics[name] = {
+                            "ap": float(box_metrics.ap[idx]),
+                            "ap50": (
+                                float(box_metrics.ap50[idx])
+                                if hasattr(box_metrics, "ap50")
+                                else 0.0
+                            ),
+                        }
+                formatted["class_metrics"] = class_metrics
+
+            return formatted
+
+        except Exception as e:
+            logger.error(f"Error formatting validation results: {e}")
+            return {"success": False, "error": str(e)}
+
+    def _format_prediction_results(self, results) -> List[Dict]:
+        """Format prediction results into list of dictionaries."""
+        detections = []
+
+        try:
+            for result in results:
+                boxes = result.boxes
+                image_path = str(result.path)
+                orig_shape = result.orig_shape  # (height, width)
+                height, width = orig_shape
+
+                # Check if this is a segmentation model with masks
+                has_masks = hasattr(result, "masks") and result.masks is not None
+
+                for i in range(len(boxes)):
+                    # Get normalized coordinates
+                    xyxyn = boxes.xyxyn[i].cpu().numpy()  # Normalized [x1, y1, x2, y2]
+
+                    detection = {
+                        "image_path": image_path,
+                        "class_id": int(boxes.cls[i]),
+                        "class_name": result.names[int(boxes.cls[i])],
+                        "confidence": float(boxes.conf[i]),
+                        "bbox_normalized": [
+                            float(v) for v in xyxyn
+                        ],  # [x_min, y_min, x_max, y_max]
+                        "bbox_absolute": [
+                            float(v) for v in boxes.xyxy[i].cpu().numpy()
+                        ],  # Absolute pixels
+                    }
+
+                    # Extract segmentation mask if available
+                    if has_masks:
+                        try:
+                            # Get the mask for this detection
+                            mask_data = result.masks.xy[
+                                i
+                            ]  # Polygon coordinates in absolute pixels
+
+                            # Convert to normalized coordinates
+                            if len(mask_data) > 0:
+                                mask_normalized = []
+                                for point in mask_data:
+                                    x_norm = float(point[0]) / width
+                                    y_norm = float(point[1]) / height
+                                    mask_normalized.append([x_norm, y_norm])
+                                detection["segmentation_mask"] = mask_normalized
+                            else:
+                                detection["segmentation_mask"] = None
+                        except Exception as mask_error:
+                            logger.warning(
+                                f"Error extracting mask for detection {i}: {mask_error}"
+                            )
+                            detection["segmentation_mask"] = None
+                    else:
+                        detection["segmentation_mask"] = None
+
+                    detections.append(detection)
+
+            return detections
+
+        except Exception as e:
+            logger.error(f"Error formatting prediction results: {e}")
+            return []
+
+    @staticmethod
+    def convert_bbox_format(
+        bbox: List[float], format_from: str = "xywh", format_to: str = "xyxy"
+    ) -> List[float]:
+        """
+        Convert bounding box between formats.
+
+        Formats:
+        - xywh: [x_center, y_center, width, height]
+        - xyxy: [x_min, y_min, x_max, y_max]
+
+        Args:
+            bbox: Bounding box coordinates
+            format_from: Source format
+            format_to: Target format
+
+        Returns:
+            Converted bounding box
+        """
+        if format_from == "xywh" and format_to == "xyxy":
+            x, y, w, h = bbox
+            return [x - w / 2, y - h / 2, x + w / 2, y + h / 2]
+        elif format_from == "xyxy" and format_to == "xywh":
+            x1, y1, x2, y2 = bbox
+            return [(x1 + x2) / 2, (y1 + y2) / 2, x2 - x1, y2 - y1]
+        else:
+            return bbox
+
+    def get_model_info(self) -> Dict[str, Any]:
+        """
+        Get information about the loaded model.
+
+        Returns:
+            Dictionary with model information
+        """
+        if self.model is None:
+            return {"error": "Model not loaded"}
+
+        try:
+            info = {
+                "model_path": self.model_path,
+                "device": self.device,
+                "task": getattr(self.model, "task", "unknown"),
+            }
+
+            # Try to get class names
+            if hasattr(self.model, "names"):
+                info["classes"] = self.model.names
+                info["num_classes"] = len(self.model.names)
+
+            return info
+
+        except Exception as e:
+            logger.error(f"Error getting model info: {e}")
+            return {"error": str(e)}

src/utils/__init__.py

@@ -0,0 +1,7 @@
+"""
+Utility modules for the microscopy object detection application.
+"""
+
+from src.utils.image import Image, ImageLoadError
+
+__all__ = ["Image", "ImageLoadError"]

src/utils/config_manager.py

@@ -0,0 +1,230 @@
+"""
+Configuration manager for the microscopy object detection application.
+Handles loading, saving, and accessing application configuration.
+"""
+
+import yaml
+from pathlib import Path
+from typing import Any, Dict, Optional
+from src.utils.logger import get_logger
+from src.utils.image import Image
+
+
+logger = get_logger(__name__)
+
+
+class ConfigManager:
+    """Manages application configuration."""
+
+    def __init__(self, config_path: str = "config/app_config.yaml"):
+        """
+        Initialize configuration manager.
+
+        Args:
+            config_path: Path to configuration file
+        """
+        self.config_path = Path(config_path)
+        self.config: Dict[str, Any] = {}
+        self._load_config()
+
+    def _load_config(self) -> None:
+        """Load configuration from YAML file."""
+        try:
+            if self.config_path.exists():
+                with open(self.config_path, "r") as f:
+                    self.config = yaml.safe_load(f) or {}
+                logger.info(f"Configuration loaded from {self.config_path}")
+            else:
+                logger.warning(f"Configuration file not found: {self.config_path}")
+                self._create_default_config()
+        except Exception as e:
+            logger.error(f"Error loading configuration: {e}")
+            self._create_default_config()
+
+    def _create_default_config(self) -> None:
+        """Create default configuration."""
+        self.config = {
+            "database": {"path": "data/detections.db"},
+            "image_repository": {
+                "base_path": "",
+                "allowed_extensions": Image.SUPPORTED_EXTENSIONS,
+            },
+            "models": {
+                "default_base_model": "yolov8s-seg.pt",
+                "models_directory": "data/models",
+                "base_model_choices": [
+                    "yolov8s-seg.pt",
+                    "yolov11s-seg.pt",
+                ],
+            },
+            "training": {
+                "default_epochs": 100,
+                "default_batch_size": 16,
+                "default_imgsz": 640,
+                "default_patience": 50,
+                "default_lr0": 0.01,
+                "two_stage": {
+                    "enabled": False,
+                    "stage1": {
+                        "epochs": 20,
+                        "lr0": 0.0005,
+                        "patience": 10,
+                        "freeze": 10,
+                    },
+                    "stage2": {
+                        "epochs": 150,
+                        "lr0": 0.0003,
+                        "patience": 30,
+                    },
+                },
+            },
+            "detection": {
+                "default_confidence": 0.25,
+                "default_iou": 0.45,
+                "max_batch_size": 100,
+            },
+            "visualization": {
+                "bbox_colors": {
+                    "organelle": "#FF6B6B",
+                    "membrane_branch": "#4ECDC4",
+                    "default": "#00FF00",
+                },
+                "bbox_thickness": 2,
+                "font_size": 12,
+            },
+            "export": {"formats": ["csv", "json", "excel"], "default_format": "csv"},
+            "logging": {
+                "level": "INFO",
+                "file": "logs/app.log",
+                "format": "%(asctime)s - %(name)s - %(levelname)s - %(message)s",
+            },
+        }
+        self.save_config()
+
+    def save_config(self) -> bool:
+        """
+        Save current configuration to file.
+
+        Returns:
+            True if successful, False otherwise
+        """
+        try:
+            # Create directory if it doesn't exist
+            self.config_path.parent.mkdir(parents=True, exist_ok=True)
+
+            with open(self.config_path, "w") as f:
+                yaml.dump(self.config, f, default_flow_style=False, sort_keys=False)
+
+            logger.info(f"Configuration saved to {self.config_path}")
+            return True
+        except Exception as e:
+            logger.error(f"Error saving configuration: {e}")
+            return False
+
+    def get(self, key: str, default: Any = None) -> Any:
+        """
+        Get configuration value by key.
+
+        Args:
+            key: Configuration key (can use dot notation, e.g., 'database.path')
+            default: Default value if key not found
+
+        Returns:
+            Configuration value or default
+        """
+        keys = key.split(".")
+        value = self.config
+
+        for k in keys:
+            if isinstance(value, dict) and k in value:
+                value = value[k]
+            else:
+                return default
+
+        return value
+
+    def set(self, key: str, value: Any) -> None:
+        """
+        Set configuration value by key.
+
+        Args:
+            key: Configuration key (can use dot notation)
+            value: Value to set
+        """
+        keys = key.split(".")
+        config = self.config
+
+        # Navigate to the nested dictionary
+        for k in keys[:-1]:
+            if k not in config:
+                config[k] = {}
+            config = config[k]
+
+        # Set the value
+        config[keys[-1]] = value
+        logger.debug(f"Configuration updated: {key} = {value}")
+
+    def get_section(self, section: str) -> Dict[str, Any]:
+        """
+        Get entire configuration section.
+
+        Args:
+            section: Section name (e.g., 'database', 'training')
+
+        Returns:
+            Dictionary with section configuration
+        """
+        return self.config.get(section, {})
+
+    def update_section(self, section: str, values: Dict[str, Any]) -> None:
+        """
+        Update entire configuration section.
+
+        Args:
+            section: Section name
+            values: Dictionary with new values
+        """
+        if section not in self.config:
+            self.config[section] = {}
+
+        self.config[section].update(values)
+        logger.debug(f"Configuration section updated: {section}")
+
+    def reload(self) -> None:
+        """Reload configuration from file."""
+        self._load_config()
+
+    def get_database_path(self) -> str:
+        """Get database path."""
+        return self.get("database.path", "data/detections.db")
+
+    def get_image_repository_path(self) -> str:
+        """Get image repository base path."""
+        return self.get("image_repository.base_path", "")
+
+    def set_image_repository_path(self, path: str) -> None:
+        """Set image repository base path."""
+        self.set("image_repository.base_path", path)
+        self.save_config()
+
+    def get_models_directory(self) -> str:
+        """Get models directory path."""
+        return self.get("models.models_directory", "data/models")
+
+    def get_default_training_params(self) -> Dict[str, Any]:
+        """Get default training parameters."""
+        return self.get_section("training")
+
+    def get_default_detection_params(self) -> Dict[str, Any]:
+        """Get default detection parameters."""
+        return self.get_section("detection")
+
+    def get_bbox_colors(self) -> Dict[str, str]:
+        """Get bounding box colors for different classes."""
+        return self.get("visualization.bbox_colors", {})
+
+    def get_allowed_extensions(self) -> list:
+        """Get list of allowed image file extensions."""
+        return self.get(
+            "image_repository.allowed_extensions", Image.SUPPORTED_EXTENSIONS
+        )

src/utils/file_utils.py

@@ -0,0 +1,239 @@
+"""
+File utility functions for the microscopy object detection application.
+"""
+
+import os
+from pathlib import Path
+from typing import List, Optional
+from src.utils.logger import get_logger
+
+
+logger = get_logger(__name__)
+
+
+def get_image_files(
+    directory: str,
+    allowed_extensions: Optional[List[str]] = None,
+    recursive: bool = False,
+) -> List[str]:
+    """
+    Get all image files in a directory.
+
+    Args:
+        directory: Directory path to search
+        allowed_extensions: List of allowed file extensions (e.g., ['.jpg', '.png'])
+        recursive: Whether to search recursively
+
+    Returns:
+        List of absolute paths to image files
+    """
+    if allowed_extensions is None:
+        from src.utils.image import Image
+
+        allowed_extensions = Image.SUPPORTED_EXTENSIONS
+
+    # Normalize extensions to lowercase
+    allowed_extensions = [ext.lower() for ext in allowed_extensions]
+
+    image_files = []
+    directory_path = Path(directory)
+
+    if not directory_path.exists():
+        logger.error(f"Directory does not exist: {directory}")
+        return image_files
+
+    try:
+        if recursive:
+            # Recursive search
+            for ext in allowed_extensions:
+                image_files.extend(directory_path.rglob(f"*{ext}"))
+                # Also search uppercase extensions
+                image_files.extend(directory_path.rglob(f"*{ext.upper()}"))
+        else:
+            # Top-level search only
+            for ext in allowed_extensions:
+                image_files.extend(directory_path.glob(f"*{ext}"))
+                # Also search uppercase extensions
+                image_files.extend(directory_path.glob(f"*{ext.upper()}"))
+
+        # Convert to absolute paths and sort
+        image_files = sorted([str(f.absolute()) for f in image_files])
+        logger.info(f"Found {len(image_files)} image files in {directory}")
+
+    except Exception as e:
+        logger.error(f"Error searching for images: {e}")
+
+    return image_files
+
+
+def ensure_directory(directory: str) -> bool:
+    """
+    Ensure a directory exists, create if it doesn't.
+
+    Args:
+        directory: Directory path
+
+    Returns:
+        True if directory exists or was created successfully
+    """
+    try:
+        Path(directory).mkdir(parents=True, exist_ok=True)
+        return True
+    except Exception as e:
+        logger.error(f"Error creating directory {directory}: {e}")
+        return False
+
+
+def get_relative_path(file_path: str, base_path: str) -> str:
+    """
+    Get relative path from base path.
+
+    Args:
+        file_path: Absolute file path
+        base_path: Base directory path
+
+    Returns:
+        Relative path string
+    """
+    try:
+        return str(Path(file_path).relative_to(base_path))
+    except ValueError:
+        # If file_path is not relative to base_path, return the filename
+        return Path(file_path).name
+
+
+def validate_file_path(file_path: str, must_exist: bool = True) -> bool:
+    """
+    Validate a file path.
+
+    Args:
+        file_path: Path to validate
+        must_exist: Whether the file must exist
+
+    Returns:
+        True if valid, False otherwise
+    """
+    path = Path(file_path)
+
+    if must_exist and not path.exists():
+        logger.error(f"File does not exist: {file_path}")
+        return False
+
+    if must_exist and not path.is_file():
+        logger.error(f"Path is not a file: {file_path}")
+        return False
+
+    return True
+
+
+def get_file_size(file_path: str) -> int:
+    """
+    Get file size in bytes.
+
+    Args:
+        file_path: Path to file
+
+    Returns:
+        File size in bytes, or 0 if error
+    """
+    try:
+        return Path(file_path).stat().st_size
+    except Exception as e:
+        logger.error(f"Error getting file size for {file_path}: {e}")
+        return 0
+
+
+def format_file_size(size_bytes: int) -> str:
+    """
+    Format file size in human-readable format.
+
+    Args:
+        size_bytes: Size in bytes
+
+    Returns:
+        Formatted string (e.g., "1.5 MB")
+    """
+    for unit in ["B", "KB", "MB", "GB"]:
+        if size_bytes < 1024.0:
+            return f"{size_bytes:.1f} {unit}"
+        size_bytes /= 1024.0
+    return f"{size_bytes:.1f} TB"
+
+
+def create_unique_filename(directory: str, base_name: str, extension: str) -> str:
+    """
+    Create a unique filename by adding a number suffix if file exists.
+
+    Args:
+        directory: Directory path
+        base_name: Base filename without extension
+        extension: File extension (with or without dot)
+
+    Returns:
+        Unique filename
+    """
+    if not extension.startswith("."):
+        extension = "." + extension
+
+    directory_path = Path(directory)
+    filename = f"{base_name}{extension}"
+    file_path = directory_path / filename
+
+    if not file_path.exists():
+        return filename
+
+    # Add number suffix
+    counter = 1
+    while True:
+        filename = f"{base_name}_{counter}{extension}"
+        file_path = directory_path / filename
+        if not file_path.exists():
+            return filename
+        counter += 1
+
+
+def is_image_file(
+    file_path: str, allowed_extensions: Optional[List[str]] = None
+) -> bool:
+    """
+    Check if a file is an image based on extension.
+
+    Args:
+        file_path: Path to file
+        allowed_extensions: List of allowed extensions
+
+    Returns:
+        True if file is an image
+    """
+    if allowed_extensions is None:
+        from src.utils.image import Image
+
+        allowed_extensions = Image.SUPPORTED_EXTENSIONS
+
+    extension = Path(file_path).suffix.lower()
+    return extension in [ext.lower() for ext in allowed_extensions]
+
+
+def safe_filename(filename: str) -> str:
+    """
+    Convert a string to a safe filename by removing/replacing invalid characters.
+
+    Args:
+        filename: Original filename
+
+    Returns:
+        Safe filename
+    """
+    # Replace invalid characters
+    invalid_chars = '<>:"/\\|?*'
+    for char in invalid_chars:
+        filename = filename.replace(char, "_")
+
+    # Remove leading/trailing spaces and dots
+    filename = filename.strip(". ")
+
+    # Ensure filename is not empty
+    if not filename:
+        filename = "unnamed"
+
+    return filename

src/utils/image.py

@@ -0,0 +1,407 @@
+"""
+Image loading and management utilities for the microscopy object detection application.
+"""
+
+import cv2
+import numpy as np
+from pathlib import Path
+from typing import Optional, Tuple, Union
+from PIL import Image as PILImage
+import tifffile
+
+from src.utils.logger import get_logger
+from src.utils.file_utils import validate_file_path, is_image_file
+
+from PySide6.QtGui import QImage
+
+logger = get_logger(__name__)
+
+
+class ImageLoadError(Exception):
+    """Exception raised when an image cannot be loaded."""
+
+    pass
+
+
+class Image:
+    """
+    A class for loading and managing images from file paths.
+
+    Supports multiple image formats: .jpg, .jpeg, .png, .tif, .tiff, .bmp
+    Provides access to image data in multiple formats (OpenCV/numpy, PIL).
+
+    Attributes:
+        path: Path to the image file
+        data: Image data as numpy array (OpenCV format, BGR)
+        pil_image: Image data as PIL Image (RGB)
+        width: Image width in pixels
+        height: Image height in pixels
+        channels: Number of color channels
+        format: Image file format
+        size_bytes: File size in bytes
+    """
+
+    SUPPORTED_EXTENSIONS = [".jpg", ".jpeg", ".png", ".tif", ".tiff", ".bmp"]
+
+    def __init__(self, image_path: Union[str, Path]):
+        """
+        Initialize an Image object by loading from a file path.
+
+        Args:
+            image_path: Path to the image file (string or Path object)
+
+        Raises:
+            ImageLoadError: If the image cannot be loaded or is invalid
+        """
+        self.path = Path(image_path)
+        self._data: Optional[np.ndarray] = None
+        self._pil_image: Optional[PILImage.Image] = None
+        self._width: int = 0
+        self._height: int = 0
+        self._channels: int = 0
+        self._format: str = ""
+        self._size_bytes: int = 0
+        self._dtype: Optional[np.dtype] = None
+
+        # Load the image
+        self._load()
+
+    def _load(self) -> None:
+        """
+        Load the image from disk.
+
+        Raises:
+            ImageLoadError: If the image cannot be loaded
+        """
+        # Validate path
+        if not validate_file_path(str(self.path), must_exist=True):
+            raise ImageLoadError(f"Invalid or non-existent file path: {self.path}")
+
+        # Check file extension
+        if not is_image_file(str(self.path), self.SUPPORTED_EXTENSIONS):
+            ext = self.path.suffix.lower()
+            raise ImageLoadError(
+                f"Unsupported image format: {ext}. "
+                f"Supported formats: {', '.join(self.SUPPORTED_EXTENSIONS)}"
+            )
+
+        try:
+            # Check if it's a TIFF file - use tifffile for better support
+            if self.path.suffix.lower() in [".tif", ".tiff"]:
+                self._data = tifffile.imread(str(self.path))
+
+                if self._data is None:
+                    raise ImageLoadError(
+                        f"Failed to load TIFF with tifffile: {self.path}"
+                    )
+
+                # Extract metadata
+                self._height, self._width = (
+                    self._data.shape[:2]
+                    if len(self._data.shape) >= 2
+                    else (self._data.shape[0], 1)
+                )
+                self._channels = (
+                    self._data.shape[2] if len(self._data.shape) == 3 else 1
+                )
+                self._format = self.path.suffix.lower().lstrip(".")
+                self._size_bytes = self.path.stat().st_size
+                self._dtype = self._data.dtype
+
+                # Load PIL version for compatibility
+                if self._channels == 1:
+                    # Grayscale
+                    self._pil_image = PILImage.fromarray(self._data)
+                else:
+                    # Multi-channel (RGB or RGBA)
+                    self._pil_image = PILImage.fromarray(self._data)
+
+                logger.info(
+                    f"Successfully loaded TIFF image: {self.path.name} "
+                    f"({self._width}x{self._height}, {self._channels} channels, "
+                    f"dtype={self._dtype}, {self._format.upper()})"
+                )
+            else:
+                # Load with OpenCV (returns BGR format) for non-TIFF images
+                self._data = cv2.imread(str(self.path), cv2.IMREAD_UNCHANGED)
+
+                if self._data is None:
+                    raise ImageLoadError(
+                        f"Failed to load image with OpenCV: {self.path}"
+                    )
+
+                # Extract metadata
+                self._height, self._width = self._data.shape[:2]
+                self._channels = (
+                    self._data.shape[2] if len(self._data.shape) == 3 else 1
+                )
+                self._format = self.path.suffix.lower().lstrip(".")
+                self._size_bytes = self.path.stat().st_size
+                self._dtype = self._data.dtype
+
+                # Load PIL version for compatibility (convert BGR to RGB)
+                if self._channels == 3:
+                    rgb_data = cv2.cvtColor(self._data, cv2.COLOR_BGR2RGB)
+                    self._pil_image = PILImage.fromarray(rgb_data)
+                elif self._channels == 4:
+                    rgba_data = cv2.cvtColor(self._data, cv2.COLOR_BGRA2RGBA)
+                    self._pil_image = PILImage.fromarray(rgba_data)
+                else:
+                    # Grayscale
+                    self._pil_image = PILImage.fromarray(self._data)
+
+                logger.info(
+                    f"Successfully loaded image: {self.path.name} "
+                    f"({self._width}x{self._height}, {self._channels} channels, "
+                    f"{self._format.upper()})"
+                )
+
+        except Exception as e:
+            logger.error(f"Error loading image {self.path}: {e}")
+            raise ImageLoadError(f"Failed to load image: {e}") from e
+
+    @property
+    def data(self) -> np.ndarray:
+        """
+        Get image data as numpy array (OpenCV format, BGR or grayscale).
+
+        Returns:
+            Image data as numpy array
+        """
+        if self._data is None:
+            raise ImageLoadError("Image data not available")
+        return self._data
+
+    @property
+    def pil_image(self) -> PILImage.Image:
+        """
+        Get image data as PIL Image (RGB or grayscale).
+
+        Returns:
+            PIL Image object
+        """
+        if self._pil_image is None:
+            raise ImageLoadError("PIL image not available")
+        return self._pil_image
+
+    @property
+    def width(self) -> int:
+        """Get image width in pixels."""
+        return self._width
+
+    @property
+    def height(self) -> int:
+        """Get image height in pixels."""
+        return self._height
+
+    @property
+    def shape(self) -> Tuple[int, int, int]:
+        """
+        Get image shape as (height, width, channels).
+
+        Returns:
+            Tuple of (height, width, channels)
+        """
+        print("shape", self._height, self._width, self._channels)
+        return (self._height, self._width, self._channels)
+
+    @property
+    def channels(self) -> int:
+        """Get number of color channels."""
+        return self._channels
+
+    @property
+    def format(self) -> str:
+        """Get image file format (e.g., 'jpg', 'png')."""
+        return self._format
+
+    @property
+    def size_bytes(self) -> int:
+        """Get file size in bytes."""
+        return self._size_bytes
+
+    @property
+    def size_mb(self) -> float:
+        """Get file size in megabytes."""
+        return self._size_bytes / (1024 * 1024)
+
+    @property
+    def dtype(self) -> np.dtype:
+        """Get the data type of the image array."""
+        if self._dtype is None:
+            raise ImageLoadError("Image dtype not available")
+        return self._dtype
+
+    @property
+    def qtimage_format(self) -> QImage.Format:
+        """
+        Get the appropriate QImage format for the image.
+
+        Returns:
+            QImage.Format enum value
+        """
+        if self._channels == 3:
+            return QImage.Format_RGB888
+        elif self._channels == 4:
+            return QImage.Format_RGBA8888
+        elif self._channels == 1:
+            if self._dtype == np.uint16:
+                return QImage.Format_Grayscale16
+            else:
+                return QImage.Format_Grayscale8
+        else:
+            raise ImageLoadError(f"Unsupported number of channels: {self._channels}")
+
+    def get_rgb(self) -> np.ndarray:
+        """
+        Get image data as RGB numpy array.
+
+        Returns:
+            Image data in RGB format as numpy array
+        """
+        if self._channels == 3:
+            return cv2.cvtColor(self._data, cv2.COLOR_BGR2RGB)
+        elif self._channels == 4:
+            return cv2.cvtColor(self._data, cv2.COLOR_BGRA2RGBA)
+        else:
+            return self._data
+
+    def get_grayscale(self) -> np.ndarray:
+        """
+        Get image as grayscale numpy array.
+
+        Returns:
+            Grayscale image as numpy array
+        """
+        if self._channels == 1:
+            return self._data
+        else:
+            return cv2.cvtColor(self._data, cv2.COLOR_BGR2GRAY)
+
+    def copy(self) -> np.ndarray:
+        """
+        Get a copy of the image data.
+
+        Returns:
+            Copy of image data as numpy array
+        """
+        return self._data.copy()
+
+    def resize(self, width: int, height: int) -> np.ndarray:
+        """
+        Resize the image to specified dimensions.
+
+        Args:
+            width: Target width in pixels
+            height: Target height in pixels
+
+        Returns:
+            Resized image as numpy array (does not modify original)
+        """
+        return cv2.resize(self._data, (width, height))
+
+    def is_grayscale(self) -> bool:
+        """
+        Check if image is grayscale.
+
+        Returns:
+            True if image is grayscale (1 channel)
+        """
+        return self._channels == 1
+
+    def is_color(self) -> bool:
+        """
+        Check if image is color.
+
+        Returns:
+            True if image has 3 or more channels
+        """
+        return self._channels >= 3
+
+    def to_normalized_float32(self) -> np.ndarray:
+        """
+        Convert image data to normalized float32 in range [0, 1].
+
+        For 16-bit images, this properly scales the full dynamic range.
+        For 8-bit images, divides by 255.
+        Already float images are clipped to [0, 1].
+
+        Returns:
+            Normalized image data as float32 numpy array [0, 1]
+        """
+        data = self._data.astype(np.float32)
+
+        if self._dtype == np.uint16:
+            # 16-bit: normalize by max value (65535)
+            data = data / 65535.0
+        elif self._dtype == np.uint8:
+            # 8-bit: normalize by 255
+            data = data / 255.0
+        elif np.issubdtype(self._dtype, np.floating):
+            # Already float, just clip to [0, 1]
+            data = np.clip(data, 0.0, 1.0)
+        else:
+            # Other integer types: use dtype info
+            if np.issubdtype(self._dtype, np.integer):
+                max_val = np.iinfo(self._dtype).max
+                data = data / float(max_val)
+            else:
+                # Unknown type: attempt min-max normalization
+                min_val = data.min()
+                max_val = data.max()
+                if max_val > min_val:
+                    data = (data - min_val) / (max_val - min_val)
+                else:
+                    data = np.zeros_like(data)
+
+        return np.clip(data, 0.0, 1.0)
+
+    def __repr__(self) -> str:
+        """String representation of the Image object."""
+        return (
+            f"Image(path='{self.path.name}', "
+            f"shape=({self._width}x{self._height}x{self._channels}), "
+            f"format={self._format}, "
+            f"size={self.size_mb:.2f}MB)"
+        )
+
+    def __str__(self) -> str:
+        """String representation of the Image object."""
+        return self.__repr__()
+
+
+def convert_grayscale_to_rgb_preserve_range(
+    pil_image: PILImage.Image,
+) -> PILImage.Image:
+    """Convert a single-channel PIL image to RGB while preserving dynamic range.
+
+    Args:
+        pil_image: Single-channel PIL image (e.g., 16-bit grayscale).
+
+    Returns:
+        PIL Image in RGB mode with intensities normalized to 0-255.
+    """
+
+    if pil_image.mode == "RGB":
+        return pil_image
+
+    grayscale = np.array(pil_image)
+    if grayscale.ndim == 3:
+        grayscale = grayscale[:, :, 0]
+
+    original_dtype = grayscale.dtype
+    grayscale = grayscale.astype(np.float32)
+
+    if grayscale.size == 0:
+        return PILImage.new("RGB", pil_image.size, color=(0, 0, 0))
+
+    if np.issubdtype(original_dtype, np.integer):
+        denom = float(max(np.iinfo(original_dtype).max, 1))
+    else:
+        max_val = float(grayscale.max())
+        denom = max(max_val, 1.0)
+
+    grayscale = np.clip(grayscale / denom, 0.0, 1.0)
+    grayscale_u8 = (grayscale * 255.0).round().astype(np.uint8)
+    rgb_arr = np.repeat(grayscale_u8[:, :, None], 3, axis=2)
+    return PILImage.fromarray(rgb_arr, mode="RGB")

src/utils/image_converters.py

@@ -0,0 +1,122 @@
+import numpy as np
+
+from roifile import ImagejRoi
+from tifffile import TiffFile, TiffWriter
+from pathlib import Path
+
+
+class UT:
+    """
+    Docstring for UT
+
+    Operetta files along with rois drawn in ImageJ
+    """
+
+    def __init__(self, roifile_fn: Path):
+        self.roifile_fn = roifile_fn
+        self.rois = ImagejRoi.fromfile(self.roifile_fn)
+        self.stem = self.roifile_fn.stem.strip("-RoiSet")
+        self.image, self.image_props = self._load_images()
+
+    def _load_images(self):
+        """Loading sequence of tif files
+        array sequence is CZYX
+        """
+        print(self.roifile_fn.parent, self.stem)
+        fns = list(self.roifile_fn.parent.glob(f"{self.stem}*.tif*"))
+        stems = [fn.stem.split(self.stem)[-1] for fn in fns]
+        n_ch = len(set([stem.split("-ch")[-1].split("t")[0] for stem in stems]))
+        n_p = len(set([stem.split("-")[0] for stem in stems]))
+        n_t = len(set([stem.split("t")[1] for stem in stems]))
+        print(n_ch, n_p, n_t)
+
+        with TiffFile(fns[0]) as tif:
+            img = tif.asarray()
+            w, h = img.shape
+            dtype = img.dtype
+            self.image_props = {
+                "channels": n_ch,
+                "planes": n_p,
+                "tiles": n_t,
+                "width": w,
+                "height": h,
+                "dtype": dtype,
+            }
+
+        image_stack = np.zeros((n_ch, n_p, w, h), dtype=dtype)
+        for fn in fns:
+            with TiffFile(fn) as tif:
+                img = tif.asarray()
+                stem = fn.stem.split(self.stem)[-1]
+                ch = int(stem.split("-ch")[-1].split("t")[0])
+                p = int(stem.split("-")[0].lstrip("p"))
+                t = int(stem.split("t")[1])
+                print(fn.stem, "ch", ch, "p", p, "t", t)
+                image_stack[ch - 1, p - 1] = img
+
+        print(image_stack.shape)
+
+        return image_stack, self.image_props
+
+    @property
+    def width(self):
+        return self.image_props["width"]
+
+    @property
+    def height(self):
+        return self.image_props["height"]
+
+    @property
+    def nchannels(self):
+        return self.image_props["channels"]
+
+    @property
+    def nplanes(self):
+        return self.image_props["planes"]
+
+    def export_rois(
+        self,
+        path: Path,
+        subfolder: str = "labels",
+        class_index: int = 0,
+    ):
+        """Export rois to a file"""
+        with open(path / subfolder / f"{self.stem}.txt", "w") as f:
+            for roi in self.rois:
+                # TODO add image coordinates normalization
+                coords = ""
+                for x, y in roi.subpixel_coordinates:
+                    coords += f"{x/self.width} {y/self.height} "
+                f.write(f"{class_index} {coords}\n")
+
+        return
+
+    def export_image(
+        self,
+        path: Path,
+        subfolder: str = "images",
+        plane_mode: str = "max projection",
+        channel: int = 0,
+    ):
+        """Export image to a file"""
+
+        if plane_mode == "max projection":
+            self.image = np.max(self.image[channel], axis=0)
+            print(self.image.shape)
+
+        with TiffWriter(path / subfolder / f"{self.stem}.tif") as tif:
+            tif.write(self.image)
+
+
+if __name__ == "__main__":
+    import argparse
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("input", type=Path)
+    parser.add_argument("output", type=Path)
+    args = parser.parse_args()
+
+    for rfn in args.input.glob("*.zip"):
+        ut = UT(rfn)
+        ut.export_rois(args.output, class_index=0)
+        ut.export_image(args.output, plane_mode="max projection", channel=0)

src/utils/logger.py

@@ -0,0 +1,75 @@
+"""
+Logging configuration for the microscopy object detection application.
+"""
+
+import logging
+import sys
+from pathlib import Path
+from typing import Optional
+
+
+def setup_logging(
+    log_file: str = "logs/app.log",
+    level: str = "INFO",
+    log_format: Optional[str] = None,
+) -> logging.Logger:
+    """
+    Setup application logging.
+
+    Args:
+        log_file: Path to log file
+        level: Logging level (DEBUG, INFO, WARNING, ERROR, CRITICAL)
+        log_format: Custom log format string
+
+    Returns:
+        Configured logger instance
+    """
+    # Create logs directory if it doesn't exist
+    log_path = Path(log_file)
+    log_path.parent.mkdir(parents=True, exist_ok=True)
+
+    # Default format if none provided
+    if log_format is None:
+        log_format = "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
+
+    # Convert level string to logging constant
+    numeric_level = getattr(logging, level.upper(), logging.INFO)
+
+    # Configure root logger
+    root_logger = logging.getLogger()
+    root_logger.setLevel(numeric_level)
+
+    # Remove existing handlers
+    root_logger.handlers.clear()
+
+    # Console handler
+    console_handler = logging.StreamHandler(sys.stdout)
+    console_handler.setLevel(numeric_level)
+    console_formatter = logging.Formatter(log_format)
+    console_handler.setFormatter(console_formatter)
+    root_logger.addHandler(console_handler)
+
+    # File handler
+    file_handler = logging.FileHandler(log_file)
+    file_handler.setLevel(numeric_level)
+    file_formatter = logging.Formatter(log_format)
+    file_handler.setFormatter(file_formatter)
+    root_logger.addHandler(file_handler)
+
+    # Log initial message
+    root_logger.info("Logging initialized")
+
+    return root_logger
+
+
+def get_logger(name: str) -> logging.Logger:
+    """
+    Get a logger instance for a specific module.
+
+    Args:
+        name: Logger name (typically __name__)
+
+    Returns:
+        Logger instance
+    """
+    return logging.getLogger(name)

src/utils/train_ultralytics_float.py

@@ -0,0 +1,561 @@
+"""
+Custom YOLO training with on-the-fly float32 conversion for 16-bit grayscale images.
+
+This module provides a custom dataset class and training function that:
+1. Load 16-bit TIFF images directly with tifffile (no PIL/cv2)
+2. Convert to float32 [0-1] on-the-fly (no data loss)
+3. Replicate grayscale to 3-channel RGB in memory
+4. Use custom training loop to bypass Ultralytics' dataset infrastructure
+5. No disk caching required
+"""
+
+import numpy as np
+import tifffile
+import torch
+from torch.utils.data import Dataset, DataLoader
+from pathlib import Path
+from typing import Optional, Dict, Any, List, Tuple
+from ultralytics import YOLO
+import yaml
+import time
+
+from src.utils.logger import get_logger
+
+logger = get_logger(__name__)
+
+
+class Float32YOLODataset(Dataset):
+    """
+    Custom PyTorch dataset for YOLO that loads 16-bit grayscale TIFFs as float32 RGB.
+
+    This dataset:
+    - Loads with tifffile (not PIL/cv2)
+    - Converts uint16 → float32 [0-1] (preserves full dynamic range)
+    - Replicates grayscale to 3 channels
+    - Returns torch tensors in (C, H, W) format
+    """
+
+    def __init__(self, images_dir: str, labels_dir: str, img_size: int = 640):
+        """
+        Initialize dataset.
+
+        Args:
+            images_dir: Directory containing images
+            labels_dir: Directory containing YOLO label files (.txt)
+            img_size: Target image size (for reference, actual resizing done by model)
+        """
+        self.images_dir = Path(images_dir)
+        self.labels_dir = Path(labels_dir)
+        self.img_size = img_size
+
+        # Find all image files
+        extensions = {".tif", ".tiff", ".png", ".jpg", ".jpeg", ".bmp"}
+        self.image_paths = sorted(
+            [
+                p
+                for p in self.images_dir.rglob("*")
+                if p.is_file() and p.suffix.lower() in extensions
+            ]
+        )
+
+        if not self.image_paths:
+            raise ValueError(f"No images found in {images_dir}")
+
+        logger.info(
+            f"Float32YOLODataset initialized with {len(self.image_paths)} images from {images_dir}"
+        )
+
+    def __len__(self):
+        return len(self.image_paths)
+
+    def _read_image(self, img_path: Path) -> np.ndarray:
+        """
+        Read image and convert to float32 [0-1] RGB.
+
+        Returns:
+            numpy array, shape (H, W, 3), dtype float32, range [0, 1]
+        """
+        # Load image with tifffile
+        img = tifffile.imread(str(img_path))
+
+        # Convert to float32
+        img = img.astype(np.float32)
+
+        # Normalize if 16-bit (values > 1.5 indicates uint16)
+        if img.max() > 1.5:
+            img = img / 65535.0
+
+        # Ensure [0, 1] range
+        img = np.clip(img, 0.0, 1.0)
+
+        # Convert grayscale to RGB
+        if img.ndim == 2:
+            # H,W → H,W,3
+            img = np.repeat(img[..., None], 3, axis=2)
+        elif img.ndim == 3 and img.shape[2] == 1:
+            # H,W,1 → H,W,3
+            img = np.repeat(img, 3, axis=2)
+
+        return img  # float32 (H, W, 3) in [0, 1]
+
+    def _parse_label(self, label_path: Path) -> List[np.ndarray]:
+        """
+        Parse YOLO label file with variable-length rows (segmentation polygons).
+
+        Returns:
+            List of numpy arrays, one per annotation
+        """
+        if not label_path.exists():
+            return []
+
+        labels = []
+        try:
+            with open(label_path, "r") as f:
+                for line in f:
+                    line = line.strip()
+                    if not line:
+                        continue
+                    # Parse space-separated values
+                    values = line.split()
+                    if len(values) >= 5:  # At minimum: class_id x y w h
+                        labels.append(
+                            np.array([float(v) for v in values], dtype=np.float32)
+                        )
+        except Exception as e:
+            logger.warning(f"Error parsing label {label_path}: {e}")
+            return []
+
+        return labels
+
+    def __getitem__(self, idx: int) -> Tuple[torch.Tensor, List[np.ndarray], str]:
+        """
+        Get a single training sample.
+
+        Returns:
+            Tuple of (image_tensor, labels, filename)
+            - image_tensor: shape (3, H, W), dtype float32, range [0, 1]
+            - labels: list of numpy arrays with YOLO format labels (variable length for segmentation)
+            - filename: image filename
+        """
+        img_path = self.image_paths[idx]
+        label_path = self.labels_dir / f"{img_path.stem}.txt"
+
+        # Load image as float32 RGB
+        img = self._read_image(img_path)
+
+        # Convert to tensor: (H, W, 3) → (3, H, W)
+        img_tensor = torch.from_numpy(img).permute(2, 0, 1).contiguous()
+
+        # Load labels (list of variable-length arrays for segmentation)
+        labels = self._parse_label(label_path)
+
+        return img_tensor, labels, img_path.name
+
+
+def collate_fn(
+    batch: List[Tuple[torch.Tensor, List[np.ndarray], str]],
+) -> Tuple[torch.Tensor, List[List[np.ndarray]], List[str]]:
+    """
+    Collate function for DataLoader.
+
+    Args:
+        batch: List of (img_tensor, labels_list, filename) tuples
+            where labels_list is a list of variable-length numpy arrays
+
+    Returns:
+        Tuple of (stacked_images, list_of_labels_lists, list_of_filenames)
+    """
+    imgs = [b[0] for b in batch]
+    labels = [b[1] for b in batch]  # Each element is a list of arrays
+    names = [b[2] for b in batch]
+
+    # Stack images - requires same H,W
+    # For different sizes, implement letterbox/resize in dataset
+    imgs_batch = torch.stack(imgs, dim=0)
+
+    return imgs_batch, labels, names
+
+
+def train_with_float32_loader(
+    model_path: str,
+    data_yaml: str,
+    epochs: int = 100,
+    imgsz: int = 640,
+    batch: int = 16,
+    patience: int = 50,
+    save_dir: str = "data/models",
+    name: str = "custom_model",
+    callbacks: Optional[Dict] = None,
+    **kwargs,
+) -> Dict[str, Any]:
+    """
+    Train YOLO model with custom Float32 dataset for 16-bit TIFF support.
+
+    Uses a custom training loop to bypass Ultralytics' dataset pipeline,
+    avoiding channel conversion issues.
+
+    Args:
+        model_path: Path to base model weights (.pt file)
+        data_yaml: Path to dataset YAML configuration
+        epochs: Number of training epochs
+        imgsz: Input image size
+        batch: Batch size
+        patience: Early stopping patience
+        save_dir: Directory to save trained model
+        name: Name for the training run
+        callbacks: Optional callback dictionary (for progress reporting)
+        **kwargs: Additional training arguments (lr0, freeze, device, etc.)
+
+    Returns:
+        Dict with training results including model paths and metrics
+    """
+    try:
+        logger.info(f"Starting Float32 custom training: {name}")
+        logger.info(
+            f"Data: {data_yaml}, Epochs: {epochs}, Batch: {batch}, ImgSz: {imgsz}"
+        )
+
+        # Parse data.yaml to get dataset paths
+        with open(data_yaml, "r") as f:
+            data_config = yaml.safe_load(f)
+
+        dataset_root = Path(data_config.get("path", Path(data_yaml).parent))
+        train_images = dataset_root / data_config.get("train", "train/images")
+        val_images = dataset_root / data_config.get("val", "val/images")
+
+        # Infer label directories
+        train_labels = train_images.parent / "labels"
+        val_labels = val_images.parent / "labels"
+
+        logger.info(f"Train images: {train_images}")
+        logger.info(f"Train labels: {train_labels}")
+        logger.info(f"Val images: {val_images}")
+        logger.info(f"Val labels: {val_labels}")
+
+        # Create datasets
+        train_dataset = Float32YOLODataset(
+            str(train_images), str(train_labels), img_size=imgsz
+        )
+        val_dataset = Float32YOLODataset(
+            str(val_images), str(val_labels), img_size=imgsz
+        )
+
+        # Create data loaders
+        train_loader = DataLoader(
+            train_dataset,
+            batch_size=batch,
+            shuffle=True,
+            num_workers=4,
+            pin_memory=True,
+            collate_fn=collate_fn,
+        )
+
+        val_loader = DataLoader(
+            val_dataset,
+            batch_size=batch,
+            shuffle=False,
+            num_workers=2,
+            pin_memory=True,
+            collate_fn=collate_fn,
+        )
+
+        # Load model
+        logger.info(f"Loading model from {model_path}")
+        ul_model = YOLO(model_path)
+
+        # Get PyTorch model
+        pt_model, loss_fn = _get_pytorch_model(ul_model)
+
+        # Setup device
+        device = kwargs.get("device", "cuda" if torch.cuda.is_available() else "cpu")
+
+        # Configure model args for loss function
+        from types import SimpleNamespace
+
+        # Required args for segmentation loss
+        required_args = {
+            "overlap_mask": True,
+            "mask_ratio": 4,
+            "task": "segment",
+            "single_cls": False,
+            "box": 7.5,
+            "cls": 0.5,
+            "dfl": 1.5,
+        }
+
+        if not hasattr(pt_model, "args"):
+            # No args - create SimpleNamespace
+            pt_model.args = SimpleNamespace(**required_args)
+        elif isinstance(pt_model.args, dict):
+            # Args is dict - MUST convert to SimpleNamespace for attribute access
+            # The loss function uses model.args.overlap_mask (attribute access)
+            merged = {**pt_model.args, **required_args}
+            pt_model.args = SimpleNamespace(**merged)
+            logger.info(
+                "Converted model.args from dict to SimpleNamespace for loss function compatibility"
+            )
+        else:
+            # Args is SimpleNamespace or other - set attributes
+            for key, value in required_args.items():
+                if not hasattr(pt_model.args, key):
+                    setattr(pt_model.args, key, value)
+
+        pt_model.to(device)
+        pt_model.train()
+
+        logger.info(f"Training on device: {device}")
+        logger.info(f"PyTorch model type: {type(pt_model)}")
+        logger.info(f"Model args configured for segmentation loss")
+
+        # Setup optimizer
+        lr0 = kwargs.get("lr0", 0.01)
+        optimizer = torch.optim.AdamW(pt_model.parameters(), lr=lr0)
+
+        # Training loop
+        save_path = Path(save_dir) / name
+        save_path.mkdir(parents=True, exist_ok=True)
+        weights_dir = save_path / "weights"
+        weights_dir.mkdir(exist_ok=True)
+
+        best_loss = float("inf")
+        patience_counter = 0
+
+        for epoch in range(epochs):
+            epoch_start = time.time()
+            running_loss = 0.0
+            num_batches = 0
+
+            logger.info(f"Epoch {epoch+1}/{epochs} starting...")
+
+            for batch_idx, (imgs, labels_list, names) in enumerate(train_loader):
+                imgs = imgs.to(device)  # (B, 3, H, W) float32
+
+                optimizer.zero_grad()
+
+                # Forward pass
+                try:
+                    preds = pt_model(imgs)
+                except Exception as e:
+                    # Try with labels
+                    preds = pt_model(imgs, labels_list)
+
+                # Compute loss
+                # For Ultralytics models, the easiest approach is to construct a batch dict
+                # and call the model in training mode which returns preds + loss
+                batch_dict = {
+                    "img": imgs,  # Already on device
+                    "batch_idx": (
+                        torch.cat(
+                            [
+                                torch.full((len(lab),), i, dtype=torch.long)
+                                for i, lab in enumerate(labels_list)
+                            ]
+                        ).to(device)
+                        if any(len(lab) > 0 for lab in labels_list)
+                        else torch.tensor([], dtype=torch.long, device=device)
+                    ),
+                    "cls": (
+                        torch.cat(
+                            [
+                                torch.from_numpy(lab[:, 0:1])
+                                for lab in labels_list
+                                if len(lab) > 0
+                            ]
+                        ).to(device)
+                        if any(len(lab) > 0 for lab in labels_list)
+                        else torch.tensor([], dtype=torch.float32, device=device)
+                    ),
+                    "bboxes": (
+                        torch.cat(
+                            [
+                                torch.from_numpy(lab[:, 1:5])
+                                for lab in labels_list
+                                if len(lab) > 0
+                            ]
+                        ).to(device)
+                        if any(len(lab) > 0 for lab in labels_list)
+                        else torch.tensor([], dtype=torch.float32, device=device)
+                    ),
+                    "ori_shape": (imgs.shape[2], imgs.shape[3]),  # H, W
+                    "resized_shape": (imgs.shape[2], imgs.shape[3]),
+                }
+
+                # Add masks if segmentation labels exist
+                if any(len(lab) > 5 for lab in labels_list if len(lab) > 0):
+                    masks = []
+                    for lab in labels_list:
+                        if len(lab) > 0 and lab.shape[1] > 5:
+                            # Has segmentation points
+                            masks.append(torch.from_numpy(lab[:, 5:]))
+                    if masks:
+                        batch_dict["masks"] = masks
+
+                # Call model loss (it will compute loss internally)
+                try:
+                    loss_output = pt_model.loss(batch_dict, preds)
+                    if isinstance(loss_output, (tuple, list)):
+                        loss = loss_output[0]
+                    else:
+                        loss = loss_output
+                except Exception as e:
+                    logger.error(f"Model loss computation failed: {e}")
+                    # Last resort: maybe preds is already a dict with 'loss'
+                    if isinstance(preds, dict) and "loss" in preds:
+                        loss = preds["loss"]
+                    else:
+                        raise RuntimeError(f"Cannot compute loss: {e}")
+
+                # Backward pass
+                loss.backward()
+                optimizer.step()
+
+                running_loss += loss.item()
+                num_batches += 1
+
+                # Report progress via callback
+                if callbacks and "on_fit_epoch_end" in callbacks:
+                    # Create a mock trainer object for callback
+                    class MockTrainer:
+                        def __init__(self, epoch):
+                            self.epoch = epoch
+                            self.loss_items = [loss.item()]
+
+                    callbacks["on_fit_epoch_end"](MockTrainer(epoch))
+
+            epoch_loss = running_loss / max(1, num_batches)
+            epoch_time = time.time() - epoch_start
+
+            logger.info(
+                f"Epoch {epoch+1}/{epochs} completed. Avg Loss: {epoch_loss:.4f}, Time: {epoch_time:.1f}s"
+            )
+
+            # Save checkpoint
+            ckpt_path = weights_dir / f"epoch{epoch+1}.pt"
+            torch.save(
+                {
+                    "epoch": epoch + 1,
+                    "model_state_dict": pt_model.state_dict(),
+                    "optimizer_state_dict": optimizer.state_dict(),
+                    "loss": epoch_loss,
+                },
+                ckpt_path,
+            )
+
+            # Save as last.pt
+            last_path = weights_dir / "last.pt"
+            torch.save(pt_model.state_dict(), last_path)
+
+            # Check for best model
+            if epoch_loss < best_loss:
+                best_loss = epoch_loss
+                patience_counter = 0
+                best_path = weights_dir / "best.pt"
+                torch.save(pt_model.state_dict(), best_path)
+                logger.info(f"New best model saved: {best_path}")
+            else:
+                patience_counter += 1
+
+            # Early stopping
+            if patience_counter >= patience:
+                logger.info(f"Early stopping triggered after {epoch+1} epochs")
+                break
+
+        logger.info("Training completed successfully")
+
+        # Format results
+        return {
+            "success": True,
+            "final_epoch": epoch + 1,
+            "metrics": {
+                "final_loss": epoch_loss,
+                "best_loss": best_loss,
+            },
+            "best_model_path": str(weights_dir / "best.pt"),
+            "last_model_path": str(weights_dir / "last.pt"),
+            "save_dir": str(save_path),
+        }
+
+    except Exception as e:
+        logger.error(f"Error during Float32 training: {e}")
+        import traceback
+
+        logger.error(traceback.format_exc())
+        raise
+
+
+def _get_pytorch_model(ul_model: YOLO) -> Tuple[torch.nn.Module, Optional[callable]]:
+    """
+    Extract PyTorch model and loss function from Ultralytics YOLO wrapper.
+
+    Args:
+        ul_model: Ultralytics YOLO model wrapper
+
+    Returns:
+        Tuple of (pytorch_model, loss_function)
+    """
+    # Try to get the underlying PyTorch model
+    candidates = []
+
+    # Direct model attribute
+    if hasattr(ul_model, "model"):
+        candidates.append(ul_model.model)
+
+    # Sometimes nested
+    if hasattr(ul_model, "model") and hasattr(ul_model.model, "model"):
+        candidates.append(ul_model.model.model)
+
+    # The wrapper itself
+    if isinstance(ul_model, torch.nn.Module):
+        candidates.append(ul_model)
+
+    # Find a valid model
+    pt_model = None
+    loss_fn = None
+
+    for candidate in candidates:
+        if candidate is None or not isinstance(candidate, torch.nn.Module):
+            continue
+
+        pt_model = candidate
+
+        # Try to find loss function
+        if hasattr(candidate, "loss") and callable(getattr(candidate, "loss")):
+            loss_fn = getattr(candidate, "loss")
+        elif hasattr(candidate, "compute_loss") and callable(
+            getattr(candidate, "compute_loss")
+        ):
+            loss_fn = getattr(candidate, "compute_loss")
+
+        break
+
+    if pt_model is None:
+        raise RuntimeError("Could not extract PyTorch model from Ultralytics wrapper")
+
+    logger.info(f"Extracted PyTorch model: {type(pt_model)}")
+    logger.info(
+        f"Loss function: {type(loss_fn) if loss_fn else 'None (will attempt fallback)'}"
+    )
+
+    return pt_model, loss_fn
+
+
+# Compatibility function (kept for backwards compatibility)
+def train_float32(model: YOLO, data_yaml: str, **train_kwargs) -> Any:
+    """
+    Train YOLO model with Float32YOLODataset (alternative API).
+
+    Args:
+        model: Initialized YOLO model instance
+        data_yaml: Path to dataset YAML
+        **train_kwargs: Training parameters
+
+    Returns:
+        Training results dict
+    """
+    return train_with_float32_loader(
+        model_path=(
+            model.model_path if hasattr(model, "model_path") else "yolov8s-seg.pt"
+        ),
+        data_yaml=data_yaml,
+        **train_kwargs,
+    )

tests/show_yolo_seg.py

@@ -0,0 +1,182 @@
+#!/usr/bin/env python3
+"""
+show_yolo_seg.py
+
+Usage:
+    python show_yolo_seg.py /path/to/image.jpg /path/to/labels.txt
+
+Supports:
+ - Segmentation polygons: "class x1 y1 x2 y2 ... xn yn"
+ - YOLO bbox lines as fallback: "class x_center y_center width height"
+Coordinates can be normalized [0..1] or absolute pixels (auto-detected).
+"""
+import sys
+import cv2
+import numpy as np
+import matplotlib.pyplot as plt
+import argparse
+from pathlib import Path
+import random
+
+
+def parse_label_line(line):
+    parts = line.strip().split()
+    if not parts:
+        return None
+    cls = int(float(parts[0]))
+    coords = [float(x) for x in parts[1:]]
+    return cls, coords
+
+
+def coords_are_normalized(coords):
+    # If every coordinate is between 0 and 1 (inclusive-ish), assume normalized
+    if not coords:
+        return False
+    return max(coords) <= 1.001
+
+
+def yolo_bbox_to_xyxy(coords, img_w, img_h):
+    # coords: [xc, yc, w, h] normalized or absolute
+    xc, yc, w, h = coords[:4]
+    if max(coords) <= 1.001:
+        xc *= img_w
+        yc *= img_h
+        w *= img_w
+        h *= img_h
+    x1 = int(round(xc - w / 2))
+    y1 = int(round(yc - h / 2))
+    x2 = int(round(xc + w / 2))
+    y2 = int(round(yc + h / 2))
+    return x1, y1, x2, y2
+
+
+def poly_to_pts(coords, img_w, img_h):
+    # coords: [x1 y1 x2 y2 ...] either normalized or absolute
+    if coords_are_normalized(coords):
+        coords = [
+            coords[i] * (img_w if i % 2 == 0 else img_h) for i in range(len(coords))
+        ]
+    pts = np.array(coords, dtype=np.int32).reshape(-1, 2)
+    return pts
+
+
+def random_color_for_class(cls):
+    random.seed(cls)  # deterministic per class
+    return tuple(int(x) for x in np.array([random.randint(0, 255) for _ in range(3)]))
+
+
+def draw_annotations(img, labels, alpha=0.4, draw_bbox_for_poly=True):
+    # img: BGR numpy array
+    overlay = img.copy()
+    h, w = img.shape[:2]
+    for cls, coords in labels:
+        if not coords:
+            continue
+        # polygon case (>=6 coordinates)
+        if len(coords) >= 6:
+            pts = poly_to_pts(coords, w, h)
+            color = random_color_for_class(cls)
+            # fill on overlay
+            cv2.fillPoly(overlay, [pts], color)
+            # outline on base image
+            cv2.polylines(img, [pts], isClosed=True, color=color, thickness=2)
+            # put class text at first point
+            x, y = int(pts[0, 0]), int(pts[0, 1]) - 6
+            cv2.putText(
+                img,
+                str(cls),
+                (x, max(6, y)),
+                cv2.FONT_HERSHEY_SIMPLEX,
+                0.6,
+                (255, 255, 255),
+                2,
+                cv2.LINE_AA,
+            )
+            if draw_bbox_for_poly:
+                x, y, w_box, h_box = cv2.boundingRect(pts)
+                cv2.rectangle(img, (x, y), (x + w_box, y + h_box), color, 1)
+        # YOLO bbox case (4 coords)
+        elif len(coords) == 4:
+            x1, y1, x2, y2 = yolo_bbox_to_xyxy(coords, w, h)
+            color = random_color_for_class(cls)
+            cv2.rectangle(img, (x1, y1), (x2, y2), color, 2)
+            cv2.putText(
+                img,
+                str(cls),
+                (x1, max(6, y1 - 4)),
+                cv2.FONT_HERSHEY_SIMPLEX,
+                0.6,
+                (255, 255, 255),
+                2,
+                cv2.LINE_AA,
+            )
+        else:
+            # Unknown / invalid format, skip
+            continue
+
+    # blend overlay for filled polygons
+    cv2.addWeighted(overlay, alpha, img, 1 - alpha, 0, img)
+    return img
+
+
+def load_labels_file(label_path):
+    labels = []
+    with open(label_path, "r") as f:
+        for raw in f:
+            line = raw.strip()
+            if not line:
+                continue
+            parsed = parse_label_line(line)
+            if parsed:
+                labels.append(parsed)
+    return labels
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Show YOLO segmentation / polygon annotations"
+    )
+    parser.add_argument("image", type=str, help="Path to image file")
+    parser.add_argument("labels", type=str, help="Path to YOLO label file (polygons)")
+    parser.add_argument(
+        "--alpha", type=float, default=0.4, help="Polygon fill alpha (0..1)"
+    )
+    parser.add_argument(
+        "--no-bbox", action="store_true", help="Don't draw bounding boxes for polygons"
+    )
+    args = parser.parse_args()
+
+    img_path = Path(args.image)
+    lbl_path = Path(args.labels)
+
+    if not img_path.exists():
+        print("Image not found:", img_path)
+        sys.exit(1)
+    if not lbl_path.exists():
+        print("Label file not found:", lbl_path)
+        sys.exit(1)
+
+    img = cv2.imread(str(img_path), cv2.IMREAD_COLOR)
+    if img is None:
+        print("Could not load image:", img_path)
+        sys.exit(1)
+
+    labels = load_labels_file(str(lbl_path))
+    if not labels:
+        print("No labels parsed from", lbl_path)
+        # continue and just show image
+    out = draw_annotations(
+        img.copy(), labels, alpha=args.alpha, draw_bbox_for_poly=(not args.no_bbox)
+    )
+
+    # Convert BGR -> RGB for matplotlib display
+    out_rgb = cv2.cvtColor(out, cv2.COLOR_BGR2RGB)
+    plt.figure(figsize=(10, 10 * out.shape[0] / out.shape[1]))
+    plt.imshow(out_rgb)
+    plt.axis("off")
+    plt.title(f"{img_path.name} ({lbl_path.name})")
+    plt.show()
+
+
+if __name__ == "__main__":
+    main()

tests/test_16bit_tiff_loading.py

@@ -0,0 +1,109 @@
+#!/usr/bin/env python3
+"""
+Test script for 16-bit TIFF loading and normalization.
+"""
+
+import numpy as np
+import tifffile
+from pathlib import Path
+import tempfile
+import sys
+import os
+
+# Add parent directory to path to import modules
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+from src.utils.image import Image
+
+
+def create_test_16bit_tiff(output_path: str) -> str:
+    """Create a test 16-bit grayscale TIFF file.
+
+    Args:
+        output_path: Path where to save the test TIFF
+
+    Returns:
+        Path to the created TIFF file
+    """
+    # Create a 16-bit grayscale test image (100x100)
+    # With values ranging from 0 to 65535 (full 16-bit range)
+    height, width = 100, 100
+
+    # Create a gradient pattern
+    test_data = np.zeros((height, width), dtype=np.uint16)
+    for i in range(height):
+        for j in range(width):
+            # Create a diagonal gradient
+            test_data[i, j] = int((i + j) / (height + width - 2) * 65535)
+
+    # Save as TIFF
+    tifffile.imwrite(output_path, test_data)
+    print(f"Created test 16-bit TIFF: {output_path}")
+    print(f"  Shape: {test_data.shape}")
+    print(f"  Dtype: {test_data.dtype}")
+    print(f"  Min value: {test_data.min()}")
+    print(f"  Max value: {test_data.max()}")
+
+    return output_path
+
+
+def test_image_loading():
+    """Test loading 16-bit TIFF with the Image class."""
+    print("\n=== Testing Image Loading ===")
+
+    # Create temporary test file
+    with tempfile.NamedTemporaryFile(suffix=".tif", delete=False) as tmp:
+        test_path = tmp.name
+
+    try:
+        # Create test image
+        create_test_16bit_tiff(test_path)
+
+        # Load with Image class
+        print("\nLoading with Image class...")
+        img = Image(test_path)
+
+        print(f"Successfully loaded image:")
+        print(f"  Width: {img.width}")
+        print(f"  Height: {img.height}")
+        print(f"  Channels: {img.channels}")
+        print(f"  Dtype: {img.dtype}")
+        print(f"  Format: {img.format}")
+
+        # Test normalization
+        print("\nTesting normalization to float32 [0-1]...")
+        normalized = img.to_normalized_float32()
+
+        print(f"Normalized image:")
+        print(f"  Shape: {normalized.shape}")
+        print(f"  Dtype: {normalized.dtype}")
+        print(f"  Min value: {normalized.min():.6f}")
+        print(f"  Max value: {normalized.max():.6f}")
+        print(f"  Mean value: {normalized.mean():.6f}")
+
+        # Verify normalization
+        assert normalized.dtype == np.float32, "Dtype should be float32"
+        assert (
+            0.0 <= normalized.min() <= normalized.max() <= 1.0
+        ), "Values should be in [0, 1]"
+
+        print("\n✓ All tests passed!")
+        return True
+
+    except Exception as e:
+        print(f"\n✗ Test failed with error: {e}")
+        import traceback
+
+        traceback.print_exc()
+        return False
+
+    finally:
+        # Cleanup
+        if os.path.exists(test_path):
+            os.remove(test_path)
+            print(f"\nCleaned up test file: {test_path}")
+
+
+if __name__ == "__main__":
+    success = test_image_loading()
+    sys.exit(0 if success else 1)

tests/test_float32_training_loader.py

@@ -0,0 +1,211 @@
+"""
+Test script for Float32 on-the-fly loading for 16-bit TIFFs.
+
+This test verifies that:
+1. Float32YOLODataset can load 16-bit TIFF files
+2. Images are converted to float32 [0-1] in memory
+3. Grayscale is replicated to 3 channels (RGB)
+4. No disk caching is used
+5. Full 16-bit precision is preserved
+"""
+
+import tempfile
+import numpy as np
+import tifffile
+from pathlib import Path
+import yaml
+
+
+def create_test_dataset():
+    """Create a minimal test dataset with 16-bit TIFF images."""
+    temp_dir = Path(tempfile.mkdtemp())
+    dataset_dir = temp_dir / "test_dataset"
+
+    # Create directory structure
+    train_images = dataset_dir / "train" / "images"
+    train_labels = dataset_dir / "train" / "labels"
+    train_images.mkdir(parents=True, exist_ok=True)
+    train_labels.mkdir(parents=True, exist_ok=True)
+
+    # Create a 16-bit TIFF test image
+    img_16bit = np.random.randint(0, 65536, (100, 100), dtype=np.uint16)
+    img_path = train_images / "test_image.tif"
+    tifffile.imwrite(str(img_path), img_16bit)
+
+    # Create a dummy label file
+    label_path = train_labels / "test_image.txt"
+    with open(label_path, "w") as f:
+        f.write("0 0.5 0.5 0.2 0.2\n")  # class_id x_center y_center width height
+
+    # Create data.yaml
+    data_yaml = {
+        "path": str(dataset_dir),
+        "train": "train/images",
+        "val": "train/images",  # Use same for val in test
+        "names": {0: "object"},
+        "nc": 1,
+    }
+    yaml_path = dataset_dir / "data.yaml"
+    with open(yaml_path, "w") as f:
+        yaml.safe_dump(data_yaml, f)
+
+    print(f"✓ Created test dataset at: {dataset_dir}")
+    print(f"  - Image: {img_path} (shape={img_16bit.shape}, dtype={img_16bit.dtype})")
+    print(f"  - Min value: {img_16bit.min()}, Max value: {img_16bit.max()}")
+    print(f"  - data.yaml: {yaml_path}")
+
+    return dataset_dir, img_path, img_16bit
+
+
+def test_float32_dataset():
+    """Test the Float32YOLODataset class directly."""
+    print("\n=== Testing Float32YOLODataset ===\n")
+
+    try:
+        from src.utils.train_ultralytics_float import Float32YOLODataset
+
+        print("✓ Successfully imported Float32YOLODataset")
+    except ImportError as e:
+        print(f"✗ Failed to import Float32YOLODataset: {e}")
+        return False
+
+    # Create test dataset
+    dataset_dir, img_path, original_img = create_test_dataset()
+
+    try:
+        # Initialize the dataset
+        print("\nInitializing Float32YOLODataset...")
+        dataset = Float32YOLODataset(
+            images_dir=str(dataset_dir / "train" / "images"),
+            labels_dir=str(dataset_dir / "train" / "labels"),
+            img_size=640,
+        )
+        print(f"✓ Float32YOLODataset initialized with {len(dataset)} images")
+
+        # Get an item
+        if len(dataset) > 0:
+            print("\nGetting first item...")
+            img_tensor, labels, filename = dataset[0]
+
+            print(f"✓ Item retrieved successfully")
+            print(f"  - Image tensor shape: {img_tensor.shape}")
+            print(f"  - Image tensor dtype: {img_tensor.dtype}")
+            print(f"  - Value range: [{img_tensor.min():.6f}, {img_tensor.max():.6f}]")
+            print(f"  - Filename: {filename}")
+            print(f"  - Labels: {len(labels)} annotations")
+            if labels:
+                print(
+                    f"  - First label shape: {labels[0].shape if len(labels) > 0 else 'N/A'}"
+                )
+
+            # Verify it's float32
+            if img_tensor.dtype == torch.float32:
+                print("✓ Correct dtype: float32")
+            else:
+                print(f"✗ Wrong dtype: {img_tensor.dtype} (expected float32)")
+                return False
+
+            # Verify it's 3-channel in correct format (C, H, W)
+            if len(img_tensor.shape) == 3 and img_tensor.shape[0] == 3:
+                print(
+                    f"✓ Correct format: (C, H, W) = {img_tensor.shape} with 3 channels"
+                )
+            else:
+                print(f"✗ Wrong shape: {img_tensor.shape} (expected (3, H, W))")
+                return False
+
+            # Verify it's in [0, 1] range
+            if 0.0 <= img_tensor.min() and img_tensor.max() <= 1.0:
+                print("✓ Values in correct range: [0, 1]")
+            else:
+                print(
+                    f"✗ Values out of range: [{img_tensor.min()}, {img_tensor.max()}]"
+                )
+                return False
+
+            # Verify precision (should have many unique values)
+            unique_values = len(torch.unique(img_tensor))
+            print(f"  - Unique values: {unique_values}")
+            if unique_values > 256:
+                print(f"✓ High precision maintained ({unique_values} > 256 levels)")
+            else:
+                print(f"⚠ Low precision: only {unique_values} unique values")
+
+            print("\n✓ All Float32YOLODataset tests passed!")
+            return True
+        else:
+            print("✗ No items in dataset")
+            return False
+
+    except Exception as e:
+        print(f"✗ Error during testing: {e}")
+        import traceback
+
+        traceback.print_exc()
+        return False
+
+
+def test_integration():
+    """Test integration with train_with_float32_loader."""
+    print("\n=== Testing Integration with train_with_float32_loader ===\n")
+
+    # Create test dataset
+    dataset_dir, img_path, original_img = create_test_dataset()
+    data_yaml = dataset_dir / "data.yaml"
+
+    print(f"\nTest dataset ready at: {data_yaml}")
+    print("\nTo test full training, run:")
+    print(f"  from src.utils.train_ultralytics_float import train_with_float32_loader")
+    print(f"  results = train_with_float32_loader(")
+    print(f"      model_path='yolov8n-seg.pt',")
+    print(f"      data_yaml='{data_yaml}',")
+    print(f"      epochs=1,")
+    print(f"      batch=1,")
+    print(f"      imgsz=640")
+    print(f"  )")
+    print("\nThis will use custom training loop with Float32YOLODataset")
+
+    return True
+
+
+def main():
+    """Run all tests."""
+    import torch  # Import here to ensure torch is available
+
+    print("=" * 70)
+    print("Float32 Training Loader Test Suite")
+    print("=" * 70)
+
+    results = []
+
+    # Test 1: Float32YOLODataset
+    results.append(("Float32YOLODataset", test_float32_dataset()))
+
+    # Test 2: Integration check
+    results.append(("Integration Check", test_integration()))
+
+    # Summary
+    print("\n" + "=" * 70)
+    print("Test Summary")
+    print("=" * 70)
+    for test_name, passed in results:
+        status = "✓ PASSED" if passed else "✗ FAILED"
+        print(f"{status}: {test_name}")
+
+    all_passed = all(passed for _, passed in results)
+    print("=" * 70)
+    if all_passed:
+        print("✓ All tests passed!")
+    else:
+        print("✗ Some tests failed")
+    print("=" * 70)
+
+    return all_passed
+
+
+if __name__ == "__main__":
+    import sys
+    import torch  # Make torch available
+
+    success = main()
+    sys.exit(0 if success else 1)

tests/test_image.py

@@ -0,0 +1,145 @@
+"""
+Tests for the Image class.
+"""
+
+import pytest
+import numpy as np
+from pathlib import Path
+from src.utils.image import Image, ImageLoadError
+
+
+class TestImage:
+    """Test cases for the Image class."""
+
+    def test_load_nonexistent_file(self):
+        """Test loading a non-existent file raises ImageLoadError."""
+        with pytest.raises(ImageLoadError):
+            Image("nonexistent_file.jpg")
+
+    def test_load_unsupported_format(self, tmp_path):
+        """Test loading an unsupported format raises ImageLoadError."""
+        # Create a dummy file with unsupported extension
+        test_file = tmp_path / "test.txt"
+        test_file.write_text("not an image")
+
+        with pytest.raises(ImageLoadError):
+            Image(test_file)
+
+    def test_supported_extensions(self):
+        """Test that supported extensions are correctly defined."""
+        expected_extensions = Image.SUPPORTED_EXTENSIONS
+        assert Image.SUPPORTED_EXTENSIONS == expected_extensions
+
+    def test_image_properties(self, tmp_path):
+        """Test image properties after loading."""
+        # Create a simple test image using numpy and cv2
+        import cv2
+
+        test_img = np.zeros((100, 200, 3), dtype=np.uint8)
+        test_img[:, :] = [255, 0, 0]  # Blue in BGR
+
+        test_file = tmp_path / "test.jpg"
+        cv2.imwrite(str(test_file), test_img)
+
+        # Load the image
+        img = Image(test_file)
+
+        # Check properties
+        assert img.width == 200
+        assert img.height == 100
+        assert img.channels == 3
+        assert img.format == "jpg"
+        assert img.shape == (100, 200, 3)
+        assert img.size_bytes > 0
+        assert img.is_color()
+        assert not img.is_grayscale()
+
+    def test_get_rgb(self, tmp_path):
+        """Test RGB conversion."""
+        import cv2
+
+        # Create BGR image
+        test_img = np.zeros((50, 50, 3), dtype=np.uint8)
+        test_img[:, :] = [255, 0, 0]  # Blue in BGR
+
+        test_file = tmp_path / "test_rgb.png"
+        cv2.imwrite(str(test_file), test_img)
+
+        img = Image(test_file)
+        rgb_data = img.get_rgb()
+
+        # RGB should have red channel at 255
+        assert rgb_data[0, 0, 0] == 0  # R
+        assert rgb_data[0, 0, 1] == 0  # G
+        assert rgb_data[0, 0, 2] == 255  # B (was BGR blue)
+
+    def test_get_grayscale(self, tmp_path):
+        """Test grayscale conversion."""
+        import cv2
+
+        test_img = np.zeros((50, 50, 3), dtype=np.uint8)
+        test_img[:, :] = [128, 128, 128]
+
+        test_file = tmp_path / "test_gray.png"
+        cv2.imwrite(str(test_file), test_img)
+
+        img = Image(test_file)
+        gray_data = img.get_grayscale()
+
+        assert len(gray_data.shape) == 2  # Should be 2D
+        assert gray_data.shape == (50, 50)
+
+    def test_copy(self, tmp_path):
+        """Test copying image data."""
+        import cv2
+
+        test_img = np.zeros((50, 50, 3), dtype=np.uint8)
+
+        test_file = tmp_path / "test_copy.png"
+        cv2.imwrite(str(test_file), test_img)
+
+        img = Image(test_file)
+        copied = img.copy()
+
+        # Modify copy
+        copied[0, 0] = [255, 255, 255]
+
+        # Original should be unchanged
+        assert not np.array_equal(img.data[0, 0], copied[0, 0])
+
+    def test_resize(self, tmp_path):
+        """Test image resizing."""
+        import cv2
+
+        test_img = np.zeros((100, 100, 3), dtype=np.uint8)
+
+        test_file = tmp_path / "test_resize.png"
+        cv2.imwrite(str(test_file), test_img)
+
+        img = Image(test_file)
+        resized = img.resize(50, 50)
+
+        assert resized.shape == (50, 50, 3)
+        # Original should be unchanged
+        assert img.width == 100
+        assert img.height == 100
+
+    def test_str_repr(self, tmp_path):
+        """Test string representation."""
+        import cv2
+
+        test_img = np.zeros((100, 200, 3), dtype=np.uint8)
+
+        test_file = tmp_path / "test_str.jpg"
+        cv2.imwrite(str(test_file), test_img)
+
+        img = Image(test_file)
+
+        str_repr = str(img)
+        assert "test_str.jpg" in str_repr
+        assert "100x200x3" in str_repr
+        assert "jpg" in str_repr
+
+        repr_str = repr(img)
+        assert "Image" in repr_str
+        assert "test_str.jpg" in repr_str

tests/test_training_dataset_prep.py

@@ -0,0 +1,142 @@
+#!/usr/bin/env python3
+"""
+Test script for training dataset preparation with 16-bit TIFFs.
+"""
+
+import numpy as np
+import tifffile
+from pathlib import Path
+import tempfile
+import sys
+import os
+import shutil
+
+# Add parent directory to path to import modules
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+from src.utils.image import Image
+
+
+def test_float32_3ch_conversion():
+    """Test conversion of 16-bit TIFF to 16-bit RGB PNG."""
+    print("\n=== Testing 16-bit RGB PNG Conversion ===")
+
+    # Create temporary directory structure
+    with tempfile.TemporaryDirectory() as tmpdir:
+        tmpdir = Path(tmpdir)
+        src_dir = tmpdir / "original"
+        dst_dir = tmpdir / "converted"
+        src_dir.mkdir()
+        dst_dir.mkdir()
+
+        # Create test 16-bit TIFF
+        test_data = np.zeros((100, 100), dtype=np.uint16)
+        for i in range(100):
+            for j in range(100):
+                test_data[i, j] = int((i + j) / 198 * 65535)
+
+        test_file = src_dir / "test_16bit.tif"
+        tifffile.imwrite(test_file, test_data)
+        print(f"Created test 16-bit TIFF: {test_file}")
+        print(f"  Shape: {test_data.shape}")
+        print(f"  Dtype: {test_data.dtype}")
+        print(f"  Range: [{test_data.min()}, {test_data.max()}]")
+
+        # Simulate the conversion process (matching training_tab.py)
+        print("\nConverting to 16-bit RGB PNG using PIL merge...")
+        img_obj = Image(test_file)
+        from PIL import Image as PILImage
+
+        # Get uint16 data
+        uint16_data = img_obj.data
+
+        # Use PIL's merge method with 'I;16' channels (proper way for 16-bit RGB)
+        if len(uint16_data.shape) == 2:
+            # Grayscale - replicate to RGB
+            r_img = PILImage.fromarray(uint16_data, mode="I;16")
+            g_img = PILImage.fromarray(uint16_data, mode="I;16")
+            b_img = PILImage.fromarray(uint16_data, mode="I;16")
+        else:
+            r_img = PILImage.fromarray(uint16_data[:, :, 0], mode="I;16")
+            g_img = PILImage.fromarray(
+                (
+                    uint16_data[:, :, 1]
+                    if uint16_data.shape[2] > 1
+                    else uint16_data[:, :, 0]
+                ),
+                mode="I;16",
+            )
+            b_img = PILImage.fromarray(
+                (
+                    uint16_data[:, :, 2]
+                    if uint16_data.shape[2] > 2
+                    else uint16_data[:, :, 0]
+                ),
+                mode="I;16",
+            )
+
+        # Merge channels into RGB
+        rgb_img = PILImage.merge("RGB", (r_img, g_img, b_img))
+
+        # Save as PNG
+        output_file = dst_dir / "test_16bit_rgb.png"
+        rgb_img.save(output_file)
+        print(f"Saved 16-bit RGB PNG: {output_file}")
+        print(f"  PIL mode after merge: {rgb_img.mode}")
+
+        # Verify the output - Load with OpenCV (as YOLO does)
+        import cv2
+
+        loaded = cv2.imread(str(output_file), cv2.IMREAD_UNCHANGED)
+        print(f"\nVerifying output (loaded with OpenCV):")
+        print(f"  Shape: {loaded.shape}")
+        print(f"  Dtype: {loaded.dtype}")
+        print(f"  Channels: {loaded.shape[2] if len(loaded.shape) == 3 else 1}")
+        print(f"  Range: [{loaded.min()}, {loaded.max()}]")
+        print(f"  Unique values: {len(np.unique(loaded[:,:,0]))}")
+
+        # Assertions
+        assert loaded.dtype == np.uint16, f"Expected uint16, got {loaded.dtype}"
+        assert loaded.shape[2] == 3, f"Expected 3 channels, got {loaded.shape[2]}"
+        assert (
+            loaded.min() >= 0 and loaded.max() <= 65535
+        ), f"Expected [0,65535] range, got [{loaded.min()}, {loaded.max()}]"
+
+        # Verify all channels are identical (replicated grayscale)
+        assert np.array_equal(
+            loaded[:, :, 0], loaded[:, :, 1]
+        ), "Channel 0 and 1 should be identical"
+        assert np.array_equal(
+            loaded[:, :, 0], loaded[:, :, 2]
+        ), "Channel 0 and 2 should be identical"
+
+        # Verify no data loss
+        unique_vals = len(np.unique(loaded[:, :, 0]))
+        print(f"\n  Precision check:")
+        print(f"    Unique values in channel: {unique_vals}")
+        print(f"    Source unique values: {len(np.unique(test_data))}")
+
+        assert unique_vals == len(
+            np.unique(test_data)
+        ), f"Expected {len(np.unique(test_data))} unique values, got {unique_vals}"
+
+        print("\n✓ All conversion tests passed!")
+        print("  - uint16 dtype preserved")
+        print("  - 3 channels created")
+        print("  - Range [0-65535] maintained")
+        print("  - No precision loss from conversion")
+        print("  - Channels properly replicated")
+
+        return True
+
+
+if __name__ == "__main__":
+    try:
+        success = test_float32_3ch_conversion()
+        sys.exit(0 if success else 1)
+    except Exception as e:
+        print(f"\n✗ Test failed with error: {e}")
+        import traceback
+
+        traceback.print_exc()
+        sys.exit(1)

tests/test_yolo_16bit_float32.py

@@ -0,0 +1,150 @@
+#!/usr/bin/env python3
+"""
+Test script for YOLO preprocessing of 16-bit TIFF images with float32 passthrough.
+Verifies that no uint8 conversion occurs and data is preserved.
+"""
+
+import numpy as np
+import tifffile
+from pathlib import Path
+import tempfile
+import sys
+import os
+
+# Add parent directory to path to import modules
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+from src.model.yolo_wrapper import YOLOWrapper
+
+
+def create_test_16bit_tiff(output_path: str) -> str:
+    """Create a test 16-bit grayscale TIFF file.
+
+    Args:
+        output_path: Path where to save the test TIFF
+
+    Returns:
+        Path to the created TIFF file
+    """
+    # Create a 16-bit grayscale test image (200x200)
+    # With specific values to test precision preservation
+    height, width = 200, 200
+
+    # Create a gradient pattern with the full 16-bit range
+    test_data = np.zeros((height, width), dtype=np.uint16)
+    for i in range(height):
+        for j in range(width):
+            # Create a diagonal gradient using full 16-bit range
+            test_data[i, j] = int((i + j) / (height + width - 2) * 65535)
+
+    # Save as TIFF
+    tifffile.imwrite(output_path, test_data)
+    print(f"Created test 16-bit TIFF: {output_path}")
+    print(f"  Shape: {test_data.shape}")
+    print(f"  Dtype: {test_data.dtype}")
+    print(f"  Min value: {test_data.min()}")
+    print(f"  Max value: {test_data.max()}")
+    print(
+        f"  Sample values: {test_data[50, 50]}, {test_data[100, 100]}, {test_data[150, 150]}"
+    )
+
+    return output_path
+
+
+def test_float32_passthrough():
+    """Test that 16-bit TIFF preprocessing passes float32 directly without uint8 conversion."""
+    print("\n=== Testing Float32 Passthrough (NO uint8) ===")
+
+    # Create temporary test file
+    with tempfile.NamedTemporaryFile(suffix=".tif", delete=False) as tmp:
+        test_path = tmp.name
+
+    try:
+        # Create test image
+        create_test_16bit_tiff(test_path)
+
+        # Create YOLOWrapper instance
+        print("\nTesting YOLOWrapper._prepare_source() for float32 passthrough...")
+        wrapper = YOLOWrapper()
+
+        # Call _prepare_source to preprocess the image
+        prepared_source, cleanup_path = wrapper._prepare_source(test_path)
+
+        print(f"\nPreprocessing result:")
+        print(f"  Original path: {test_path}")
+        print(f"  Prepared source type: {type(prepared_source)}")
+
+        # Verify it returns a numpy array (not a file path)
+        if isinstance(prepared_source, np.ndarray):
+            print(
+                f"\n✓ SUCCESS: Prepared source is a numpy array (float32 passthrough)"
+            )
+            print(f"  Shape: {prepared_source.shape}")
+            print(f"  Dtype: {prepared_source.dtype}")
+            print(f"  Min value: {prepared_source.min():.6f}")
+            print(f"  Max value: {prepared_source.max():.6f}")
+            print(f"  Mean value: {prepared_source.mean():.6f}")
+
+            # Verify it's float32 in [0, 1] range
+            assert (
+                prepared_source.dtype == np.float32
+            ), f"Expected float32, got {prepared_source.dtype}"
+            assert (
+                0.0 <= prepared_source.min() <= prepared_source.max() <= 1.0
+            ), f"Expected values in [0, 1], got [{prepared_source.min()}, {prepared_source.max()}]"
+
+            # Verify it has 3 channels (RGB)
+            assert (
+                prepared_source.shape[2] == 3
+            ), f"Expected 3 channels (RGB), got {prepared_source.shape[2]}"
+
+            # Verify no quantization to 256 levels (would happen with uint8 conversion)
+            unique_values = len(np.unique(prepared_source))
+            print(f"  Unique values: {unique_values}")
+
+            # With float32, we should have much more than 256 unique values
+            if unique_values > 256:
+                print(f"\n✓ SUCCESS: Data has {unique_values} unique values (> 256)")
+                print(f"  This confirms NO uint8 quantization occurred!")
+            else:
+                print(f"\n✗ WARNING: Data has only {unique_values} unique values")
+                print(f"  This might indicate uint8 quantization happened")
+
+            # Sample some values to show precision
+            print(f"\n  Sample normalized values:")
+            print(f"    [50, 50]: {prepared_source[50, 50, 0]:.8f}")
+            print(f"    [100, 100]: {prepared_source[100, 100, 0]:.8f}")
+            print(f"    [150, 150]: {prepared_source[150, 150, 0]:.8f}")
+
+            # No cleanup needed since we returned array directly
+            assert (
+                cleanup_path is None
+            ), "Cleanup path should be None for float32 pass through"
+
+            print("\n✓ All float32 passthrough tests passed!")
+            return True
+
+        else:
+            print(f"\n✗ FAILED: Prepared source is a file path: {prepared_source}")
+            print(f"  This means data was saved to disk, not passed as float32 array")
+            if cleanup_path and os.path.exists(cleanup_path):
+                os.remove(cleanup_path)
+            return False
+
+    except Exception as e:
+        print(f"\n✗ Test failed with error: {e}")
+        import traceback
+
+        traceback.print_exc()
+        return False
+
+    finally:
+        # Cleanup
+        if os.path.exists(test_path):
+            os.remove(test_path)
+            print(f"\nCleaned up test file: {test_path}")
+
+
+if __name__ == "__main__":
+    success = test_float32_passthrough()
+    sys.exit(0 if success else 1)

tests/test_yolo_16bit_preprocessing.py

@@ -0,0 +1,126 @@
+#!/usr/bin/env python3
+"""
+Test script for YOLO preprocessing of 16-bit TIFF images.
+"""
+
+import numpy as np
+import tifffile
+from pathlib import Path
+import tempfile
+import sys
+import os
+
+# Add parent directory to path to import modules
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+from src.model.yolo_wrapper import YOLOWrapper
+from src.utils.image import Image
+from PIL import Image as PILImage
+
+
+def create_test_16bit_tiff(output_path: str) -> str:
+    """Create a test 16-bit grayscale TIFF file.
+
+    Args:
+        output_path: Path where to save the test TIFF
+
+    Returns:
+        Path to the created TIFF file
+    """
+    # Create a 16-bit grayscale test image (200x200)
+    # With values ranging from 0 to 65535 (full 16-bit range)
+    height, width = 200, 200
+
+    # Create a gradient pattern
+    test_data = np.zeros((height, width), dtype=np.uint16)
+    for i in range(height):
+        for j in range(width):
+            # Create a diagonal gradient
+            test_data[i, j] = int((i + j) / (height + width - 2) * 65535)
+
+    # Save as TIFF
+    tifffile.imwrite(output_path, test_data)
+    print(f"Created test 16-bit TIFF: {output_path}")
+    print(f"  Shape: {test_data.shape}")
+    print(f"  Dtype: {test_data.dtype}")
+    print(f"  Min value: {test_data.min()}")
+    print(f"  Max value: {test_data.max()}")
+
+    return output_path
+
+
+def test_yolo_preprocessing():
+    """Test YOLO preprocessing of 16-bit TIFF images."""
+    print("\n=== Testing YOLO Preprocessing of 16-bit TIFF ===")
+
+    # Create temporary test file
+    with tempfile.NamedTemporaryFile(suffix=".tif", delete=False) as tmp:
+        test_path = tmp.name
+
+    try:
+        # Create test image
+        create_test_16bit_tiff(test_path)
+
+        # Create YOLOWrapper instance (no actual model loading needed for this test)
+        print("\nTesting YOLOWrapper._prepare_source()...")
+        wrapper = YOLOWrapper()
+
+        # Call _prepare_source to preprocess the image
+        prepared_path, cleanup_path = wrapper._prepare_source(test_path)
+
+        print(f"\nPreprocessing complete:")
+        print(f"  Original path: {test_path}")
+        print(f"  Prepared path: {prepared_path}")
+        print(f"  Cleanup path: {cleanup_path}")
+
+        # Verify the prepared image exists
+        assert os.path.exists(prepared_path), "Prepared image should exist"
+
+        # Load the prepared image and verify it's uint8 RGB
+        prepared_img = PILImage.open(prepared_path)
+        print(f"\nPrepared image properties:")
+        print(f"  Mode: {prepared_img.mode}")
+        print(f"  Size: {prepared_img.size}")
+        print(f"  Format: {prepared_img.format}")
+
+        # Convert to numpy to check values
+        img_array = np.array(prepared_img)
+        print(f"  Shape: {img_array.shape}")
+        print(f"  Dtype: {img_array.dtype}")
+        print(f"  Min value: {img_array.min()}")
+        print(f"  Max value: {img_array.max()}")
+        print(f"  Mean value: {img_array.mean():.2f}")
+
+        # Verify it's RGB uint8
+        assert prepared_img.mode == "RGB", "Prepared image should be RGB"
+        assert img_array.dtype == np.uint8, "Prepared image should be uint8"
+        assert img_array.shape[2] == 3, "Prepared image should have 3 channels"
+        assert (
+            0 <= img_array.min() <= img_array.max() <= 255
+        ), "Values should be in [0, 255]"
+
+        # Cleanup prepared file if needed
+        if cleanup_path and os.path.exists(cleanup_path):
+            os.remove(cleanup_path)
+            print(f"\nCleaned up prepared image: {cleanup_path}")
+
+        print("\n✓ All YOLO preprocessing tests passed!")
+        return True
+
+    except Exception as e:
+        print(f"\n✗ Test failed with error: {e}")
+        import traceback
+
+        traceback.print_exc()
+        return False
+
+    finally:
+        # Cleanup
+        if os.path.exists(test_path):
+            os.remove(test_path)
+            print(f"Cleaned up test file: {test_path}")
+
+
+if __name__ == "__main__":
+    success = test_yolo_preprocessing()
+    sys.exit(0 if success else 1)