It is very challenging to accurately identify small objects in remote sensing images.In order to solve the problem of small object detection in remote sensing images,this paper proposes an improved YOLOv5.Firstly,Mosaic-9 is used to preprocess the data set to solve the problem of data scarcity.Secondly,the CA attention mechanism module is added to the backbone to in-crease the perception of detailed information and improve the detection ability of small objects.Finally,BiFPN is introduced into the feature fusion network to solve the problem that the object features are few and easy to be lost.The experimental results show that mAP of the improved YOLOv5 for small object detection in remote sensing images reaches 90.3%,which is superior to the conven-tional object detection model and suitable for small object detection in remote sensing images.
维普
万方数据