Unmanned aerial vehicle multi-target long-term tracking method

The invention provides an unmanned aerial vehicle multi-target long-term tracking method, and the method comprises the steps that a detection module recognizes target detection frame images and target relative position information of all monitoring targets from environment image information, and obtains a target detection frame image sequence and a target relative position information sequence; the cross-mirror tracking module extracts multi-dimensional apparent characteristics of each target detection frame image; and for each frame in the target detection frame image sequence, after the matching module performs anti-shake operation on the frame, the frame is divided into a high-score target frame and a low-score target frame according to confidence, a fusion loss matrix is obtained based on multi-dimensional apparent characteristics and target relative position information, and then track matching is performed through a Hungary algorithm to complete target tracking. According to the invention, the lens anti