A Lightweight Improved Yolov5 for Insulator Defect Detection
Xinping Zhu · 2023
The safety of insulators is related to the stability of power transmission and economic livelihood. UAV inspection has the importance of improving detection efficiency and adapting to complex terrain. In this paper, based on MobileNetV3 and coordinate attention, the YOLOv5 detection algorithm is improved. The number of parameters is reduced by 2.1 GFLOPs compared with the YOLOv5n model, and the ability of the network to learn the relative positions of defects and insulators is strengthened, so that the detection accuracy increases. It has the potential to be applied in UAV transmission line inspection.