An Improved YOLOv3 Algorithm and Intruder Detection on Transmission Line

For detecting foreign bodies in transmission lines, the manual inspection method has low accuracy and a certain degree of risk. Therefore, this paper proposes an improved YOLOv3-based transmission line foreign body detection method for intelligent inspection. The accuracy of the improved YOLOv3 model for transmission line foreign body detection is obviously raised. Through ablation experiments, it verifies the effectiveness of the foreign body detection method proposed in this paper. Compared with other advanced target detection algorithms, the proposed algorithm has achieved a well-balanced accuracy and speed of detection, and more competitive for transmission line foreign body detection tasks.