Abstract: Abnormal heating frequently occurs before outdoor cable terminations fail. Timely and accurate diagnosis therefore requires intelligent aerial inspection and real-time diagnosis. Key challenges include automatically identifying heat-prone areas, aligning the camera with the target, adaptively focusing the lens, detecting obstacles and generating avoidance prompts, extracting temperature data, and diagnosing defects. This paper develops an intelligent aerial inspection and real-time diagnosis system for abnormal heating defects of outdoor cable terminations. Based on object detection, edge detection, and depth estimation algorithms, the system completes nine principal steps to diagnose abnormal heating defects accurately and intelligently in real time. It substantially reduces dependence on personnel experience and improves the operational reliability of cable lines.