Radar and Image Fusion for Power Line Detection in UAV Applications

Power lines are conceivable obstacles for UAV applications, such as door-to-door delivery services. This paper proposes radar and image fusion for power line detections as part of the work to implement a UAV autopilot system. It employs an mm-wave automotive radar, and its parameters are optimized for power line detection. The image detection is utilized with YOLO for real-time objection detection, and its detection processes are modified to obtain a better recognition precision rate. Both radar and image detection systems are integrated under the LabVIEW environment. The experiments demonstrated that radar and image fusion enhances the power line detection rate.