Improved Remote Sensing Image Target Detection Algorithm Based on YOLOv5

Remote sensing information processing plays an important role in many fields such as environmental protection, urban planning, military reconnaissance and so on. However, there are problems in remote sensing images such as complex background, large size change, small target, etc. A remote sensing image target detection algorithm to improve YOLOv5 is proposed. Using a certain percentage of data enhancement technology at the input side to enrich the data features and improve the model generalization ability and robustness; for the problem of small targets in remote sensing images, SPD-Conv module is added to reduce the loss of fine-grained information and improve the detection performance of small targets; the design uses the DCNv2 module to replace part of the neck structure of the C3 layer, which breaks through the limitation of traditional convolution and improves the feature extraction ability of the network in remote sensing images. Experimental results show that the improved YOLOv5 model improves the detection accuracy by 3.9%.