AI-Driven Overspeeding Detection 🚗💨

Overview

This project is an AI-powered system for detecting overspeeding vehicles using computer vision. It was developed as part of the PES Hackathon. The system identifies vehicles in a video, detects their speed, and marks license plates, highlighting those exceeding the speed limit (50 km/h).

Features

• 📷 Real-time Vehicle Detection using a pre-trained AI model (YOLOv8 and YOLOv11).
• 🏎 Speed Estimation to determine if a vehicle is overspeeding.
• 🔍 License Plate Recognition for identifying vehicles.
• 🚨 Automatic Overspeed Highlighting for vehicles exceeding 50 km/h.
• 📊 Video Output with Annotations showing detected vehicles, their speed, and license plates.
• 🌐 Web Dashboard for RTO officials to monitor violations.

Tech Stack

• Python 🐍
• OpenCV 🎥
• YOLO / Custom AI Model 🤖
• OCR (Tesseract / EasyOCR) 🔠
• Node.js (Backend API) 🌐
• React.js (Frontend) ⚛️
• Google Colab (For Model Training) 💻
• Kaggle (For Dataset) 📊

Installation

1️⃣ Clone the Repository

git clone https://github.com/Pranava-Pai-N/AI-DRIVEN-OVERSPEEDING-DETECTION-PES-HACKATHON.git
cd AI-DRIVEN-OVERSPEEDING-DETECTION-PES-HACKATHON

2️⃣ Install Dependencies

Frontend Setup (React + Vite)

cd frontend
npm install  # Install frontend dependencies

Backend Setup (Node.js + Express)

cd backend
npm install  # Install backend dependencies

Python Dependencies (For Detection & ANPR)

If your project includes a Python-based AI model, install the required dependencies:

pip install -r requirements.txt

---

Usage

1️⃣ Run the Backend Server

cd backend
node server.js

The backend will start on http://localhost:5000.

2️⃣ Run the Frontend Dashboard

cd frontend
npm run dev

The frontend will start on http://localhost:5173.

3️⃣ Run ANPR and Detection Script

Run the AI-based detection script with a video input:

cd ml
python anpr.py

Expected Output

• A processed video with marked license plates.
• Speed displayed on-screen for each detected vehicle.
• Vehicles exceeding 50 km/h highlighted.
• A web dashboard for RTO officials providing details about speed violations.

Future Enhancements

• 🚀 Real-time processing on live camera feeds.
• 📌 Geo-location tagging for violations.
• 📡 Cloud integration for centralized monitoring.
• 📱 Mobile App Integration for Real-time Monitoring.

Contributors

• Pranav Pai N
• Ajith Goveas
• Aadithya Nayak V
• Karthik Acharya

License

This project is open-source and available under the MIT License.

---

📫 Feel free to reach out or contribute! 🚀