Abstract: In this study, a method for overhead high-voltage wire segmentation and inspection is proposed to address the difficulty and low efficiency of overhead high-voltage cable inspection. This method utilizes the histogram bimodal method, region of interest extraction, and a wire optimization method that combines filters and image differentiation to extract and eliminate interference items, such as clouds, towers, and the ground in the image. The improved Hough line detection algorithm, which optimizes the voting and screening mechanisms, extracts the wires, and finally intercepts the required part of the wires through the topological relationship between the insulators and wires. The experimental results show that the proposed algorithm has ideal detection performance, with an average intersection-to-parallel ratio of 94.4% and a wire detection accuracy of 92.8%, meeting actual industrial production requirements.