Abstract:

Recognition of helmetless motorcycles is essential for everyone’s safety. Riding a motorcycle without a safety helmet is a criminal offense as well as one of the leading causes of accidents. To address this, a unique method of identifying riders without helmets has been developed. This method uses deep learning technology to identify riders wearing helmets and displays the rider's registration number. where the YOLO method is used in the model design to detect objects. The primary item in our model is the helmet, which is identified, and then the optical character recognition technique is used to display the registration plate of the rider without a helmet. We create a website that allows users to log in using their credentials. The webpage will then appear with an image upload button. In order to recognize a helmet, the model first determines whether a rider and a vehicle are present. The status of helmet detection is then shown. The vehicle's registration number is extracted if there is no helmet using OCR.

Description:

Deep learning, YOLO, OCR, COCO, Flask, SQLite