Abstract:In order to improve the identification rate of external hidden damage targets under complex backgrounds, a fusion approach combining deep learning and millimeter-wave radar is adopted. and it combines the high-precision distance information from millimeter-wave radar with the rich environmental information captured by monitoring camera. Through sensor calibration, time synchronization and spatial integration between millimeter-wave radar and monitoring camera are achieved. Based on the YOLOv5 target identification algorithm, the images captured by the monitoring camera are processed in-depth to accurately identify the target objects in the images and output corresponding detection frames and results. Meanwhile, the target points detected by the millimeter-wave radar are mapped to the images to generate regions of interest, enabling the algorithm to focus more on these areas that may contain external hidden damage, thus improving the efficiency and accuracy of detection. Finally, through the target fusion algorithm and combining the identification and detection results of monitoring camera and millimeter-wave radar, the external hidden damage targets of the power transmission line are precisely identified and classified, effectively solving the impact of light, distance and other factors on traditional monitoring devices, and achieving precise identification of external hidden damage targets of the power transmission line under complex backgrounds.