HELMET RULE VIOLATION DETECTION USING DEEP LEARNING

The widespread non-compliance of helmet-wearing regulations among motorcyclists in India poses a significant threat to road safety. This paper introduces a new approach to address this issue through the developing a Helmet Rule Violation Detection system. Given the challenges faced by traffic police in enforcing helmet regulations, the proposedsystem leverages computer vision techniques to automatically identify motorcyclists without helmets. The methodology involves the detection of moving vehicles through thresholding, followed by a classification process based on area and aspect ratio to distinguish between motorcyclists and non-motorcyclists. The subsequent identification of helmet violation is achieved by isolating relevant blobs and employing a classification mechanism. By automating this process, the system aims to reduce the burden on traffic police and enhance the overall safety of road users. The proposed approach not only contributes to the mitigation of helmet rule violations but also promotes responsible road behavior, fostering a safer and more compliant traffic environment in India.