A Method for Detecting Insulator Defects in Unmanned Aerial Vehicle Aerial Photography Based on Improved YOLOv5

The background of insulator defects in transmission lines captured by drones is complex, and existing deep learning algorithms have low detection accuracy and usually require significant computational resources. To address this issue, we propose an improved YOLOv5 insulator defect detection algorithm in this paper. First, we introduce the GhostConv structure, and a G-C3 module is further designed to reduce the computational complexity. Then, we utilize a mobile attention mechanism and build a CA-S attention fusion module to improve the feature extraction capability of the network. Experimental results show that compared with the traditional YOLOv5 model, the average detection accuracy of the improved YOLOv5 proposed in the paper is increased by 1.9%, and the number of parameters is reduced by 12.5%. It is proved that the proposed algorithm has better performance than other popular algorithms in insulator defect detection of transmission lines.