Helmet Detection on Two-Wheeler Riders with Deep Learning

Helmet Detection on Two-Wheeler Riders with Deep Learning

Shrishailappa Patil, Madhura Pande, Sahish Pandav, Soham Panchal, Shivam Pandagale, Yogesh Pande, Ghanshyam Pandey

Department of Engineering, Sciences and Humanities (DESH)

Vishwakarma Institute of Technology, Pune, 411037, Maharashtra, India

Abstract — This project focuses on building a software that determines whether a motorcyclist is wearing a helmet using Machine Learning (ML). The system is able to detect this through video capture. It hopes to help in detection of riders without a helmet. The system is successfully able to detect helmets with a probability of 80%. It makes use of OpenCV, TensorFlow, and YOLOv5.

It recognizes the number plate of the bike if its rider is without a helmet and captures its cropped image and saves it on the local storage of the system.

Keywords —Haar-cascade, Helmet detection, Machine Learning, Number plate capturing, OpenCV, Python, TensorFlow, YOLOv5