Autonomous Vision-Based UAV Landing with Collision Avoidance Using Deep Learning

Abstract

AbstractThe autonomous vision-based Unmanned Aerial Vehicles (UAVs) landing is an adaptive way to land in special environments such as the global positioning system denied. There is a risk of collision when multiple UAVs land simultaneously without communication on the same platform. This work accomplishes vision-based autonomous landing and uses a deep-learning-based method to realize collision avoidance during the landing process. Specifically, the landing UAVs are categorized into Level I and II. The YoloV4 deep learning method will be implemented by the Level II UAV to achieve object detection of Level I UAV. Once the Level I UAV’s landing has been detected by the onboard camera of Level II UAV, it will move and land on a relative landing zone beside the Level I UAV. The experiment results show the validity and practicality of our theory. KeywordsVision-basedCollision avoidanceDeep learning