Improved YOLOv5 Traffic Sign Detection

Abstract

Aiming at the problems such as low accuracy of traffic sign detection and poor real-time performance, a traffic sign detection algorithm based on YOLOv5 is proposed. First, the C2f module is introduced into the backbone network to obtain richer gradient flow information and enhance the feature fusion capability of the target. Second, the SimAM attention module is introduced into the backbone network to enhance the target features and weaken the background features to improve the feature extraction capability of the network model, which in turn improves the detection accuracy of the network model. The experimental results show that compared with the original algorithm, the mAP@0.5 increased by 1.5%, the mAP@0.5:0.95 increased by 2.6%, and the detection speed increased by 28.82%. The improved algorithm detection accuracy can reach 95.5%, and the detection speed can reach 58.82FPS, which can meet the requirements of real-time accurate detection of traffic sign detection.