Yolov5-Based System for Detecting Helmet Violations and Overloading in Two-Wheelers

The Two-Wheeler Violation Prediction System is a real-time object detection framework designed toenhance road safety by monitoring helmet usage and the number of riders on motorcycles. It employs theYOLOv5 model to detect violations, such as missing helmets and overloaded vehicles, from video feeds andwebcams. Trained on a custom dataset annotated with Roboflow, the system ensures accuracy through imagepreprocessing and enhancement. It includes a helmet detection model to assess compliance with traffic laws.Additionally, it features license plate recognition using Roboflow’s API and TesseractOCR, storing violationdata in an Excel database for enforcement. The system can save frames of violations for documentation andlegal use. Designed for scalability, it integrates with traffic monitoring, urban surveillance, and lawenforcement checkpoints, automating violation detection to improve traffic safety and law enforcementefficiency.