Motorcyclist safety is a critical concern for traffic officials, especially regarding the mandatory use of helmets in accident prevention. However, during the recent epidemic, there has been a noticeable trend where individuals prioritize wearing masks over helmets, potentially leading to safety hazards and traffic violations. To address this issue, we embarked on a project aimed at identifying and penalizing helmetless riders for traffic regulation violations. Additionally, we integrated number plate recognition using the AWS Rekognition service, leveraging a Faster R-CNN model for object detection in images. AWS Rekognition detects helmets by generating bounding boxes around them, which provide spatial and proportional data relative to the rider's bounding box. These separate entities enable the system to determine helmet usage accurately. Our project proposes using image or video footage to automatically detect motorcyclists without helmets and capture motorbike license plate numbers using Easy OCR.

Motorcyclist safety, helmet detection, traffic regulation enforcement, AWS Rekognition, Faster R-CNN, object detection, number plate recognition