DESIGN AND IMPLEMENTATION OF HELMET DETECTION

Motorcycle safety helmets are used to prevent head injuries while riding bike. In many countries, due to the lack of police power to successfully implement helmet laws, many bike riders don’t wear helmets. This paper presents an Automated helmet detection system which automatically detects whether motorcycle riders are wearing safety helmets or not. Number of motorcycle accidents have been rapidly increasing throughout recent years in many countries. Motorcycle is becoming increasingly popular because of various social and economic factors. The helmet is the very important safety equipment for motorcyclists but for some reasons they do not want to use it. An accident can be fatal if motorcyclist isn’t wearing helmet because it causes head injury. For image processing, at first we detect vehicles that are moving in real-time by doing background extraction from the image. Background Subtraction is used for the extraction process. After that the resulting image is enhanced using peak and morphological method. Feature extraction is done using appropriate methods and Neural Network is used for image classification. At last, a helmet is detected using the Hough Transform method.