Intelligent Accident Detection System using Machine Learning

Road accidents continue to be a major threat to public safety, with delayed reporting being one of the primary reasons for increased fatalities. This work introduces an AI-driven Accident Detection and Emergency Alert System that analyzes real-time traffic footage using a custom Convolutional Neural Network (CNN). The system automatically detects accident-like patterns from live camera feeds and immediately passes the alert to a Flask-based backend. The backend generates a structured report, stores details in MongoDB, and sends automated notifications to emergency responders. A Next.js dashboard displays live alerts, images, timestamps, and severity levels for administrative review. The integrated approach ensures quick detection, rapid reporting, and reduced human involvement, contributing to more efficient emergency management.