Research on Improved Algorithm for Safety Helmet Detection Based on YOLOv5
Tang Zhi-hang, Yi Liang He, Tang Jiale, Zhao Si Meng · International Journal of Advanced Networking and Applications · 2025
In order to improve the accuracy and robustness of helmet wearing detection, the algorithm was further optimized based on YOLOv5, combined with CBAM attention mechanism, BiFPN structure, Focal loss function and other technologies, significantly improving the accuracy and precision of helmet wearing detection and recognition, comprehensively enhancing the level of safety protection in the production process, and meeting the accuracy and real-time requirements of helmet wearing detection. The experimental results show that the improved helmet detection algorithm has a 3.2% increase in mAP compared to the original YOLOv5 algorithm.