Helmet detection algorithm based on improved yolov7

Abstract

In order to solve the problems of low detection accuracy, wrong detection and missed detection in the traditional target detection algorithm in the safety helmet detection task, a safety helmet detection algorithm based on improved YOLOv7 is proposed. The algorithm is based on the YOLOv7 framework structure. Firstly, in the backbone network, the 3x3 convolution in the ELAN structure is replaced by the SAC convolution, and the receptive field size is adaptively selected in combination with the dilated convolution to improve the detection accuracy. At the same time, the SimAM attention mechanism is introduced to focus on key information and improve the accuracy of model detection. Finally, WIoU loss is used to replace the cross entropy in the original loss function to improve the weight of positive and negative samples, accelerate the convergence speed of the model, and further improve the detection accuracy. The experimental results show that the accuracy of the algorithm on the safety helmet data set is 3.6 percentage points higher than that of YOLOv7, which improves the detection accuracy of the safety helmet and has certain application value.