1. Rider helmet detection using YOLOv5 image targeting algorithm.
- Author
-
Sharma, Bhavishya, Barve, Nachiketa, and Ramalingam, Anita
- Subjects
- *
HELMETS , *TRAFFIC police , *BUILDING sites , *ALGORITHMS , *MOTORCYCLISTS - Abstract
Two-wheeler vehicles are a major form of personal transport in the world and specially India, a substantial number of 2 wheelers on roads have made it difficult to monitor them correctly. This has led to carelessness from the drivers and increase in the number of accidents that can be prevented and minimized. One of the protective gears-helmet is not always worn by riders and it is tough for the traffic police to catch these violators, due to sheer size. Our project involves using the YOLO (You Only Look Once) algorithm to detect whether the riders as well as pillions are wearing the helmets so that they can be caught, and their vehicle details be sent for further processing to the traffic police. Along with the 2 wheelers, helmets are also required at construction sites, and we plan to apply the same. This paper proposes a framework for identification of motorcyclists without helmets. The application, with the help of YOLOv5 (You Only Look Once) and rigorous training gives us an accuracy of 87% true positives detection. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF