A Monocular Ranging Method Based on YOLOv8 for UAV Formation

Complex environments such as mountainous areas and cities often interfere with GPS and communication, making it difficult for drone clusters to maintain formation and affecting work quality. We propose a drone monocular ranging method based on YOLOv8-Pose, which first uses a deep-learning network model to achieve drone detection and reference points positioning, and then uses the PnP algorithm to calculate the distance. This method balances the speed and accuracy of ranging and can help drones maintain formation in scenarios where GPS and commu-nication are disrupted.