YOLO-Powered Detection of Helmets and Number Plates for Enforcing Safety on Wheels

Road safety is becoming an increasingly important issue, particularly in a country like India, where two-wheelers are a common mode of transport. A major concern is that many riders do not wear helmets, which can lead to serious injuries or even death in the event of an accident. This paper proposes an intelligent system that automatically recognizes riders not wearing helmets and saves their images and vehicle number plates using a combination of computer vision and deep learning. This system uses the YOLO object detection algorithm to locate riders, helmets, and number plates in real-time video streams. A custom-trained optical character recognition model is then used to extract and recognize the characters on the plate. All recognized number plates are stored in a CSV file, from which we can send fines to the corresponding riders.