The fixed-point landing of the quad-rotor UAV based on YOLOv5

This study proposed a method for realizing a fixed- point landing of a quad-rotor UAV based on YOLOv5. The experimental equipment contained a quad-rotor UAV and two edge-computing devices, and an intelligent connected vehicle (ICV) was used as the landing platform; the data transmission between experimental equipment was realized by local area network technology through a Wi-Fi module. The quad-rotor UAV was developed independently, including the UAV frame and flight control module, co-processing module, airborne camera and gimbal stabilizer, GPS locator, barometer, and power supply system. Real-time target detection was performed on the UAV's shooting pictures on an edge-computing device, and the detection results were sent to another edge-computing device. Another edge- computing device controlled the movement of the UAV according to the detection result, to realize the fixed-point landing of the quad-rotor UAV. Finally, the experimental results showed that the YOLOv5 target detection method can meet the real-time and accuracy requirements of the quad-rotor UAV for fixed-point landing.