In order to improve the multi-scale pedestrian detection performance in various scenes,an improved multi-scale YOLO-V5 algorithm combined with attention mechanism is proposed.By deepening the YOLO-V5 backbone network,the feature extraction ability is further improved and deep semantic information is enriched;the Coordinate Attention atten-tion mechanism is introduced into YOLO-V5 to focus on the effective area of the input fea-ture map;a new prediction head is added to the original YOLO-V5 to enhance its detection performance for small targets.The proposed method has been tested on the Citypersons dataset and after subdividing its pedestrian targets in validation set into three different scales,the AP50 values for three different scales pedestrian targets reached 64.5%,66.6%,71.7%respectively and the Recall values reached 53.0%,56.6%and 61.7%respectively,which were 3.8%,3.6%,2.3%and 3.3%,4.7%,3.5%higher than the original YOLO-V5.The experimental results show that the proposed algorithm can obviously improve the multi-scale pedestrian detection performance.
维普
万方数据