Abstract: In the maintenance and operation of power systems, it is crucial to detect foreign objects hanging on transmission lines, especially for high-voltage overhead transmission lines. A device mounted on unmanned aerial vehicle for high-voltage inspection drone loading device was designed, using the YOLOv5 (You only look once version 5) algorithm with a self-constructed foreign object data set to train the model, and combining various sensors, to achieve timely and effective detection of foreign objects on high-voltage transmission lines or high-voltage transmission towers. The feasibility of this design was validated through model testing. The device connects to the unmanned aerial vehicle via a type-C interface as an external load, ensuring no safety hazards and expanding the application scope of the unmanned aerial vehicle with minimal hardware requirements.