Seat Belt And Helmet Detection Using Deep Learning

The proposed system relates the current situation of multitudinous problems in traffic rules and regulations in different countries. Further, examines how any accident or causality can be averted using enhance methods of natural language processing. Mainly in highways, people just focus on speed and overtaking the other vehicles, rather than following traffic rules and taking basic safety measures. And due to these highways, police barricading is also not possible. Therefore, it is highly requisite to adopt an automatic seat belt and helmet detection system which captures the image of rule brakers, so that necessary legal action should be taken against them. The system works on real time detection and used for continuous surveillance for safety purposes. In the paper we solve the issue of manually checking the safety measures while driving a vehicle to avoid any accidents and deaths due to fast moving vehicles without wearing proper safety measures. The elucidated process in the paper is based on real-time surveillance system and obligate everyone to wear seat belts and helmets to increase the safety of individuals by the help of OpenCV (Open-Source Computer Vision Library) and deep learning algorithm which is embedded in Raspberry Pi. Addition to this, all the pictures and videos are saved in a database which is only accessed by higher government officials for future security purpose and expediate the tackling process of traffic rules violation due to the negligence and irresponsibility of law enforcement officials as well as common citizens.