Helmet Detection and Triple Ride Detection using CNN and YOLO

Motorcycles have been the primary means of trans- portation in developing countries. Unfortunately, the motorcycle accident cases have been very high recently. Most of the death toll in these motorcycle accidents is due to the fact that motorcyclists fail to wear helmets. How best the motorcyclists are ensured to put on their helmets goes a long way with which technology is utilized: from monitoring the police manual at the intersections or CCTV footage to capture those without a helmet. However these methods consume considerable human effort and interaction. This system contemplates an automated scheme for the detection of helmet less motorcyclists and retrieves license number from CCTV footage. The system first categorizes moving objects as motorcycles or otherwise. For classified motorcyclists, the system determines whether they wear helmets or not. In case, the motorcyclist does not wear a helmet, the system retrieves the License plate number using an OCR algorithm. In this project CNN and yolo algorithms are used for recognition of person with and without helmet as well as triple riding detection. The violated person number plate is detected if its visible. Keywords: Motorcycles, Helmet usage, CCTV footage, automated approach, Non-helmeted motorcyclists, license plate, moving ob- jects, CNN (Convolutional Neural Networks), YOLO (You Only Look Once), recognition, OCR (Optical Character Recognition)