Defect detection of transmission lines based on SH-YOLOv8

Power transmission lines are an extremely important component of the power system. With the rapid development of China's power industry, the inspection tasks of power transmission lines face significant challenges, among which the detection of insulator defects, faulty cables, and tower corrosion is a crucial part of the inspection process. In response to the issue of power transmission line defect detection, an improved YOLOv8 algorithm is proposed. The SH-YOLOv8 model, based on the structure of the YOLOv8 algorithm, integrates a Cross-Scale Feature Aggregation Module (CAFM) to further enhance the capability of detecting defects in power transmission lines, especially in complex backgrounds and multi-scale object detection. The introduction of the CAFM module allows SH-YOLOv8 to more efficiently integrate features of different scales, achieving stronger cross-level information sharing. Experimental results demonstrate that the improved SH-YOLOv8 performs exceptionally well in the task of power transmission line defect detection, particularly in the recognition of small-scale targets such as missing insulators and faulty cables, significantly reducing the incidence of missed and false detections. This provides an efficient and accurate technical solution for the intelligent development of power grid inspection.