AI Traffic Violation Detector

A computer-vision–based system designed to monitor traffic violations in real time, addressing challenges such as signal jumping, helmetless riding, overspeeding, wrong-lane driving, and triple-riding.

Features

• Real-time Detection: Analyzes live CCTV feeds for multiple violation types
• Object Detection: Uses YOLOv8 for detecting vehicles, riders, and traffic signals
• Object Tracking: DeepSORT for tracking vehicles and riders across frames
• Violation Detection: Specialized logic for each violation type
• Evidence Generation: Timestamped images and video clips with license plate data
• Privacy Protection: Face blurring for privacy-by-design
• Dashboard: Streamlit-based interface for reviewing violations
• Edge Deployment: Optimized for near-real-time performance

Installation

1. Clone or download the project
2. Create a virtual environment:

   python -m venv venv

3. Activate the virtual environment:
   • Windows: venv\Scripts\activate
   • Linux/Mac: source venv/bin/activate
4. Install dependencies:

   pip install -r requirements.txt

Usage

Running the Detector

python main.py

This will start live detection using your default webcam. To use a video file:

from main import TrafficViolationDetector
detector = TrafficViolationDetector()
detector.run_live_detection('path/to/video.mp4')

Running the Dashboard

streamlit run dashboard.py

Access the dashboard at http://localhost:8501

Project Structure

traffic_violation_detector/
├── config.py              # Configuration settings
├── detection.py           # Object detection modules
├── tracking.py            # Object tracking with DeepSORT
├── violations.py          # Violation detection logic
├── evidence.py            # Evidence generation and management
├── main.py                # Main application
├── dashboard.py           # Streamlit dashboard
├── requirements.txt       # Python dependencies
├── README.md              # This file
├── models/                # Pre-trained models (to be added)
├── evidence/              # Generated evidence storage
│   ├── images/
│   └── clips/
└── logs/                  # Application logs

Violation Types Detected

1. Signal Jumping: Vehicles crossing stop lines during red lights
2. Helmetless Riding: Motorcyclists without helmets
3. Overspeeding: Vehicles exceeding speed limits
4. Wrong-lane Driving: Vehicles in incorrect lanes
5. Triple Riding: More than 2 persons on motorcycles

Configuration

Edit config.py to customize:

• Model paths
• Detection thresholds
• Video settings
• Evidence storage locations
• Privacy settings

Requirements

• Python 3.8+
• Webcam or CCTV feed
• Tesseract OCR (for license plate recognition)
• Sufficient storage for evidence

Future Enhancements

• Custom helmet detection model
• Lane detection for wrong-lane violations
• Traffic light color recognition
• Integration with traffic authority systems
• Mobile app for officers
• Advanced analytics and reporting

License

This project is for educational and research purposes. Ensure compliance with local laws and privacy regulations when deploying.

Contributing

Contributions are welcome! Please open issues for bugs or feature requests.