Helmet Detection Using Machine Learning and Automatic Number Plate Recognition

Abstract— Traffic safety is critically compromised when motorcyclists ride without helmets, increasing the risk of serious injuries and fatalities. Manual monitoring of helmet compliance is inefficient and prone to errors, highlighting the need for automation. This project introduces an automated system for detecting helmet usage using machine learning techniques. The system utilizes the YOLO v3 (You Only Look Once) object detection algorithm, designed to identify helmet use among motorcyclists in real-time. By integrating advanced image processing and YOLO v3, the system ensures high accuracy and swift detection. This real-time monitoring system can be employed to assist law enforcement agencies in promoting helmet use and improving road safety. Keywords— Helmet Detection, YOLO v3, Machine Learning, Convolutional Neural Networks (CNN), Road Safety, Traffic Surveillance.