Target Standoff Tracking Method Based on Aerial Photography Information from Small UAVs

Aiming at the problem of poor tracking accuracy or even tracking failure due to irregular motion and scale change of targets during tracking and striking ground maneuvering targets by small UAVs, a cascading target standoff tracking framework based on aerial photography information is constructed. An active target localization algorithm based on YOLOv8n is proposed, which realizes accurate detection and fast localization of small targets on the ground. The target state and observation equations are established, and for the nonlinear characteristics of the equations, a Cubature Kalman Filtering (CKF) algorithm is used to predict the target motion state. The Lyapunov Vector Field (LVF) is introduced to construct a globally convergent limit loop as a standoff tracking desired path, and the target speed estimated by the filter is corrected to the UAV speed to realize the stable standoff tracking of the UAV to the maneuvering target. The experimental results show that this method has excellent performance in maneuvering target positioning accuracy and tracking stability, and can effectively enhance the reconnaissance and surveillance capability of UAVs and provide more accurate support for target strikes, which is of great practical value.