Integrating computer vision technology into a video inspection system enables image processing and object analysis of video data streams to identify the status of power supply equipment, such as the positions of switch disconnectors, signal indicators, pressure plates, meters, and protection device panels - replacing manual judgment with machine recognition to ultimately achieve unmanned inspections. After intelligently identifying the image-based status, the results are compared with those of the PSCADA system to achieve "dual-system verification" of the equipment status. An alarm is triggered when the information from the two sources is inconsistent.