Abstract: In the operation and management of external insulation of electrical equipment in substations, the external insulation status is usually detected by electrical detection method, which is susceptible to interference from external electromagnetic fields, resulting in low accuracy (ACC) values of the detection results. Therefore, a power outage detection method for external insulation of electrical equipment in substations based on ultraviolet imaging technology is proposed. Relying on ultraviolet imaging technology, on-site observation images of external insulation components are collected in uninterrupted scenarios, and median filtering is performed to improve the quality of ultraviolet image data, which serves as the basis for subsequent external insulation state detection. Using the C-V model to perform UV image segmentation, observing information such as the area and perimeter of the target area, and extracting image features. Establish a feature fusion convolutional neural network model, deeply learn and classify input feature quantities, and obtain intelligent detection results for external insulation without power outage. From the experimental results, it can be seen that the ACC value provided by this method reaches 0.99, which can accurately reflect the working status of the external insulation components of electrical equipment.