This paper presents a laser cleaning vision inspection system, leveraging computer vision technology, designed to intelligently detect and manage rust stains on the metal surfaces of outdoor disconnect switches in substations. The system employs the HSV color conversion method to establish color thresholds based on rust color characteristics, facilitating the initial extraction of rust-affected areas. Subsequent processing involves Gaussian filtering and morphological image operations—expansion and erosion—to accurately identify rust regions and ascertain their location and dimensions. This information is then fed into a pulse laser controller to execute the laser cleaning of rusted areas on the workpiece surface. The system was implemented in a substation in Ulanqab, demonstrating effective rust removal from the metal surfaces of the contact finger workpieces of the disconnect switches. The system achieved a rust area recognition accuracy of 93.10% under the HSV color space, thereby accomplishing the design of a machine vision-based laser cleaning system for rust on disconnect switches.