Research on Paper Bag Defects Detection Based on Improved YOLOv5 Algorithm
In order to improve the detection precision of paper bag manufacturers on the handle or bottom de-fects,a paper bag defects detection method based on the improved YOLOv5 algorithm was proposed.In order to improve the network positioning ability and enhance the feature learning and expression ability of the net-work,the Coordinate Attention was introduced,and then the EIoU loss function was introduced to improve the loss function,so as to improve the rationality of the aspect ratio of the original network loss function to im-prove the regression precision,and a simplified spatial pyramid pooling structure with CSP-like structure was introduced to reduce redundant information processing and improve the detection performance.Experimental results show that mAP@.5 and mAP@.5∶.95 of the improved algorithm are 87.3%and 56.8%,respective-ly.Compared with the YOLOv5 algorithm,mAP@.5 and mAP@.5∶.95 of the improved algorithm are in-creased by 1.6%and 0.9%respectively,showing better performance in the detection of paper bag defects.
paper bag defectsattentionloss functionspatial pyramid poolingimproved algorithm
NETL
NSTL
万方数据
