Small Target Traffic Sign Detection Algorithm Based on Improved YOLOv7
In automatic driving,the problem of detecting the accuracy and leakage of small targets is cru-cial for the accuracy of vehicle equipment in judging road conditions.Aiming at the problem of small traf-fic sign detection accuracy and missed detection,a small target traffic sign detection algorithm based on improved YOLOv7 is proposed.The ACmix attention module is introduced into Neck,which is used to improve the sensitivity of the network to small-scale targets and reduce the impact of noise;the multi-head self-attention mechanism SPD convolutional building block is introduced at the output side to im-prove the detection performance of small traffic sign targets at the output side;and the loss function is op-timized by using SIoU to replace the CIoU in the original YOLOv7 network that improves the robustness of the network.To verify the performance of the proposed algorithm,experimental validation is carried out on the CCTSDB traffic sign dataset.The experimental results show that for the small targets in the data-set,the improved YOLOv7 network,compared with the original YOLOv7 network,the leakage detection is significantly improved and the map reaches 98.7％,and the operation speed reaches 110.5 f/s,which is an improvement in both accuracy and speed compared with the original YOLOv7 model,and it can sat-isfy the requirements of traffic sign detection.
万方数据
维普
