AI Traffic Enforcement

About This Project

The AI Traffic Enforcement System is a full-scale, smart traffic monitoring solution built using YOLOv8, OpenCV, and a hybrid Flask + Django backend to automate the detection and processing of traffic violations. Designed for the Nirman Hackathon’24, this system combines real-time computer vision, ANPR (Automatic Number Plate Recognition), and automated communication workflows to streamline traffic law enforcement. The system detects vehicles through a YOLOv8-based model, tracks their movement, measures speed, and identifies overspeeding incidents. It then uses Haar Cascade / EasyOCR for accurate number-plate extraction and validation. Once a violation is confirmed, the backend automatically generates a PDF memo using FPDF, attaches violation details and images, and sends notifications via Twilio SMS and Gmail email service, including a short link for downloading the memo. An integrated admin panel enables staff and authorities to log in using secure OTP-based authentication, view all generated memos, manage user and vehicle data, filter records by speed/date/road, and track historical violations. The system stores all records securely in Firebase for real-time access and analytics. This project showcases expertise in AI-driven traffic monitoring, real-time video processing, OCR, backend automation, PDF generation, and full-stack dashboard development, offering a complete end-to-end automated solution for modern smart-city traffic enforcement.