Lavideol/Vehicle-Detection-Nepal

Vehicle Detection and Classification System - Nepal

A comprehensive vehicle detection and classification system built for Nepali roads using advanced computer vision techniques.

Features

• —Real-time Vehicle Detection: Detects vehicles in images and videos using YOLO
• —Vehicle Classification: Classifies vehicles by make and model
• —Multi-format Support: Works with images (JPG, PNG) and videos (MP4, WebM, etc.)
• —Detailed Analytics: Provides statistics and visualizations of detected vehicles
• —Web Interface: User-friendly Flask-based dashboard

Supported Vehicle Classes

Bikes

• —Bajaj, Hero, Honda, Jawa, KTM, Mahindra, Royal Enfield, Suzuki, TVS, Yamaha

Scooters

• —Aprilia, Hero Destini, Honda Activa, Honda Dio, Suzuki Access, TVS Jupiter, TVS NTorq, Vespa, Yadea, Yamaha Ray

Cars

• —Ford, Honda, Hyundai, Kia, MG, Nissan, Suzuki, Tata, Toyota, Volkswagen

Commercial Vehicles

• —Bus, Truck, Hiace

How to Use

1. 1.Upload Content: Upload an image or video file using the dashboard
2. 2.View Results: See detected vehicles with bounding boxes and classifications
3. 3.Download Reports: Get CSV reports with detailed vehicle information

Technologies Used

• —Framework: Flask
• —Detection: YOLOv8 (Ultralytics)
• —Classification: Custom YOLO classifier
• —Computer Vision: OpenCV
• —Data Processing: Pandas

Installation & Running Locally

# Install dependencies
pip install -r requirements.txt

# Run the application
python app.py

The app will be available at http://localhost:7860

Model Files

Models are automatically downloaded on first run:

• —YOLOv8n for vehicle detection
• —Custom classifier for vehicle type classification

Docker Deployment

docker build -t vehicle-detection-nepal .
docker run -p 7860:7860 vehicle-detection-nepal

Results

Results are saved as CSV files with:

• —Vehicle type and make/model
• —Detection confidence scores
• —Frame timestamps (for videos)
• —Bounding box coordinates

License

This project is designed for educational and commercial use.

Credits

Built as a semester project for vehicle detection and classification in Nepal.