A comprehensive system for detecting, analyzing, and indexing road defects using YOLOv10 computer vision and fuzzy logic severity assessment.
- Real-time Detection - YOLOv10-powered defect detection via camera or video file
- Multi-Defect Support - Linear cracks, alligator cracks, and potholes
- GPS Integration - Precise location tracking for all detected defects
- Cloud Storage - Automated data backup and synchronization
- Live Dashboard - Real-time visualization with statistics and analysis
- Camera Controls - Zoom, flip, and multi-camera support
- Web Mapping - Interactive defect visualization
- Camera: 720p minimum (1280x720+ recommended)
- GPS module: NMEA-compatible (required)
- GPU: NVIDIA with CUDA support (recommended)
- Python 3.8+
- OpenCV 4.x
- PyQt5
- CUDA Toolkit (for GPU acceleration)
- Dependencies in
requirements.txt
- Clone the repository
git clone https://github.com/yourusername/Road-Defect-Indexing-System.git
cd Road-Defect-Indexing-System- Run automated setup
setup.batAutomatically handles: Python verification, virtual environment creation, package installation, CUDA detection, directory setup, and model verification.
- Launch application
start.batpython -m src.appRoad-Defect-Indexing-System/
├── src/
│ ├── app.py # Application entry point
│ ├── modules/
│ │ ├── camera.py # Camera handling
│ │ ├── detection.py # Defect detection engine
│ │ ├── gps_reader.py # GPS integration
│ │ └── cloud_connector.py # Cloud storage
│ ├── ui/ # GUI components
│ │ ├── dashboard.py
│ │ ├── video_controls.py
│ │ ├── main_controls.py
│ │ ├── statistics.py
│ │ └── status_bar.py
│ └── models/
│ ├── yolov10/ # YOLOv10 submodule
│ └── model.pt # Trained model
├── public/icons/
├── scripts/testscripts/
├── requirements.txt
├── setup.bat
└── start.bat
1. Video Source
- Camera: Select from available devices
- Video File: Upload for batch processing
2. Camera Controls (Live mode)
- Zoom adjustment
- View orientation flip
3. Playback Controls (Video mode)
- Play/Pause, Rewind/Forward
- Progress tracking
4. Detection
- Start/Stop detection
- Real-time results visualization
- Live statistics monitoring
5. GPS & Cloud
- Auto-connects to GPS module
- Coordinates logged per defect
- Cloud sync for data backup
6. Analysis (Coming Soon)
- Detailed defect reports
- Severity trend analysis
- Export capabilities
Access via Settings dialog:
- Output directories
- Recording options
- Cloud storage credentials
- Analysis parameters
Detection Pipeline
- Image acquisition
- Preprocessing
- YOLOv10 detection
- Post-processing
- Fuzzy logic severity calculation
GPS Integration
- NMEA protocol support
- Automatic port detection
- Real-time coordinate logging
Cloud Storage
- Efficient data transmission
- Metadata and image backup
| Issue | Solution |
|---|---|
| Camera not detected | Check connections, verify permissions, try different camera index |
| Low detection accuracy | Improve lighting, verify camera focus |
| GPS connection failed | Check COM port, ensure clear sky view |
| Performance lag | Enable GPU acceleration, reduce resolution, close background apps |
Road Defect Mapping Web App
Interactive Mapbox dashboard for visualizing detected defects with GPS precision and road-type filtering. Part of the Comprehensive Road Defect Indexing System.
- Fork the repository
- Create feature branch (
git checkout -b feature/AmazingFeature) - Commit changes (
git commit -m 'Add AmazingFeature') - Push to branch (
git push origin feature/AmazingFeature) - Open Pull Request
MIT License - see LICENSE file for details.
- YOLO team for detection framework
- OpenCV and PyQt communities
- All contributors and maintainers
v1.0.0 - Initial release with core detection, dashboard, GPS integration, and cloud storage support.
Support: Open an issue for questions or bug reports.