Third Eye: A Comprehensive Solution for Road Safety

Abstract

Factors such as reckless and careless driving have led to a rapid increase in two-wheeler accidents, leading to an increased fatality rate. As per a recent United Nations survey, lives of at least four in ten motorcyclists who die in road accidents could have been saved by wearing proper head safety gear. It becomes utmost clear to impose strict helmet regulations. However, the limited presence of traffic police personnel makes it impossible to pursue legal action against every violator. Thus, automating this process of handling violations is highly desirable. We propose a system that will monitor live feeds from inputs such as traffic surveillance systems, vehicle dashcams, and traffic personnel bodycams to detect the offense. The system in discussion will be lightweight and optimized to work efficiently on existing commodity hardware. Suppose the motorcyclist is not wearing a helmet. In that case, the system will record the offense by capturing the license plate information along with snapshots of the vehicle, location and timestamp of the offense.