Low-altitude Airspace Conflict Detection and Avoidance Algorithm Based on Computer Vision and Multi-sensors

The development and application bring a large quantity of unmanned aerial vehicle (UAV) flight activities, which directly affect the flight safety in complex airspace. UAV path planning algorithm is based on the task type to be completed, and then combined with the known planning space and terrain information, it plans an optimal path for UAV or UAV group that conforms to the real flying of UAV. Improving UAV’s perception and avoidance ability and avoiding the collision danger between UAVs or between UAVs and manned aircraft is a core ability for UAV to enter non-isolated airspace. In order to prevent aircraft from colliding in low-altitude airspace, this paper analyzes the conflict prediction and avoidance methods between aircraft flying at the same altitude in low-altitude airspace, and puts forward the conflict detection and avoidance algorithm in low-altitude airspace based on binocular stereo vision and multi-sensors. In order to get more accurate position information faster, stereo vision and optical flow are combined. The experimental results show that this method can effectively improve the efficiency and accuracy of avoidance.