Safety Helmet Detection Using Improved YOLOv8

The wearing of safety helmet directly affects the life safety of workers in the construction site or production workshop, and it is of great significance for the supervision of workers’ helmet wearing. In this paper, for the current helmet wearing detection algorithm in the complex environment detection accuracy is not high, put forward a yolov8-based improved helmet detection algorithm YOLOv8-Focaler-SE, in the network structure to add SE attention mechanism module, so that the network more accurately locate the region of interest in the target, at the same time in the CIOU loss function on the basis of the introduction of the Focaler-CIoU loss, thus thereby thereby improving the overall performance of the algorithm. Experimental results on the safety helmet wearing dataset show that YOLOv8-Focaler-SE improves the precision, recall, mAP 50 and mAP $50-95$ metrics by $1.2 \%$, $2.1 \%, 1.5 \%$, and $2.1 \%$, respectively, while the number of model parameters remains essentially unchanged when compared to the yolov8n algorithm. The method improves the performance of the detection algorithm, which can accurately detect helmet wearing and provide technical support for safety production.