Traffic accidents are one of the most common causes of death in the world, and helmet use has been effective proved in reducing the risk of head injuries for motorcyclists. Therefore, it is crucial to ensure that motorcyclists always wear helmets while riding. One method to detect helmet use is by utilizing object recognition technology based on Convolutional Neural Networks (CNN). This study focus on design and implement a simple helmet detection system using CNN methods and the YOLO v3 model for real-time detection. The system is expected to accurately detect helmet use by riders. In this research, the YOLO v3 model is trained using the COCO dataset, which includes various images with diverse contexts. The results of this implementation show that the system can detect helmet use effectively under various lighting conditions and environments. This demonstrates the potential use of a helmet detection system based on CNN and YOLO in enhancing riding safety.
Copyrights © 2024