A Deep Learning-Based Smart Helmet Detection and Notification with e-Challan Issuance for Traffic Fines

The Ministry of Road Transport and Highways estimates that 158,964 two-wheeler road accidents occurred in India in 2020, resulting in 56,873 fatalities, some of which were brought on by the lack of a helmet. This is still an issue on Indian roads and puts the lives of the country's young inhabitants in peril, despite numerous campaigns that have been undertaken by various state governments and the imposition of numerous punishments. The usage of computer vision is a suitable remedy for this issue. An automated system is suggested that employs a state-of-the-art single shot detector (SSD) for detection, image upscaling and Optical Character Recognition (OCR) to identify the vehicle's licence plate, as well as a database for e-challan records. Together with the payment link to clear the challan online, other Short Message Service (SMS) and mail Application Programming Interface (API) is also utilised to notify about the challan's issuance. This paper explores a deep-learning-based idea that assures traffic safety and order on Indian roads and is a workable and useful solution to the highlighted issue.