An improved UAV image detection algorithm based on YOLOv5

At present, existing small target detection methods suffer from low detection accuracy, high false detection rate and high leakage rate. In order to improve small target detection, this paper proposes the FSD-YOLOv5 algorithm, which has three improvements on the basis of the YOLOv5 algorithm. Firstly, the Focal EIoU is used instead of the original CIoU to improve the model convergence speed and regression accuracy; Secondly, in order to cope with the defects in the CNN architecture, we adopt a new CNN building block called SPD-Conv; Finally, to address the problem of the reduced or lost information of small objects in feature maps caused by downsampling in convolutional neural networks, we introduce feature reuse to increase the feature information of small objects in the feature maps. We evaluated FSD-YOLOv5 on the VisDrone-2019 dataset and compared its performance with YOLOv5. The results show that FSD-YOLOv5 achieves a detection accuracy of 36.3%, an improvement of 2.4% compared to the original algorithm. For small target detection, the FSD-YOLOv5 algorithm has better detection results than the YOLOv5 algorithm.