Semantic Segmentation and Localization of Doublesplit Diversion Line Based on Improved Yolov8

In order to make the diversion line accurately penetrate into the grounding wire clamp hole position through the actuation structure without human intervention, the computer must accurately acquire the position of the diversion line relative to the end of the operation. Images captured in real substation scenes have a large number of interference backgrounds similar to diversion lines, and the foreground and background areas of the images are not clearly differentiated, which greatly increases the difficulty of semantic segmentation and localization of double-split diversion lines. Aiming at this problem, this paper designs a semantic segmentation and localization method for double-split diversion lines in complex scenarios of substations.