Implementation of YOLOV7 to Recognize Traffic Regulation Infractions

Abstract

In the realm of motorcycle safety, helmets are paramount for rider protection, as mandated by regulations in many countries. Unfortunately, some riders disregard these safety measures by either opting not to wear helmets or wearing them incorrectly. Computer vision technologies have emerged as a key tool in designing intelligent traffic systems to address this issue. These technologies leverage techniques like Image detection in the background and foreground to recognize moving objects in a scene and extract relevant features. Machine learning algorithms are then employed to categorize these detected items, contributing to enhanced road safety and compliance with helmet regulations. Our project takes advantage of YOLO V7, a powerful deep learning model designed for object detection. This model aids us in the identification and classification of bikers without helmets or bikes carrying more than two passengers, ensuring strict adherence to traffic regulations for safe motorcycle riding. Moreover, our system features a user-friendly frontend user interface, simplifying user interaction and facilitating ease of use.