Improved YOLOv5s Method for Yarn Tube Object Detection
To address the issues of low efficiency and high errors in traditional yarn tube classification methods,a yarn tube object recognition method based on an improved YOLOv5s algorithm is proposed.The enhanced YOLOv5s algorithm integrates a coordinate attention(CA)module and a Transformer mod-ule,introduces a new spatial pyramid pooling plus(SPP+)module to replace the conventional spatial pyr-amid pooling(SPP)module,enhances feature fusion using the weighted bidirectional feature pyramid net-work(BiFPN)concept,and replaces the original loss function with the wise intersection over Union(WIoU)loss function.To validate the performance of the improved algorithm,a yarn tube dataset is crea-ted,and yarn tube detection experiments are conducted based on the improved YOLOv5s algorithm.The experimental results show that the improved algorithm exhibits superior recognition capabilities.
万方数据
维普
