Positioning method research for unmanned aerial vehicles based on Meanshift tracking algorithm

摘要

The positioning algorithm of unmanned aerial vehicle (UAV) is one of the key technologies in the field of unmanned aerial vehicle autonomous flight. In this paper, the absolute position of unmanned aerial vehicle is calculated without global positioning system (GPS) signals by using moving targets tracked by onboard camera and the relative position between it and UAV. The positioning method is presented by combined the Meanshift target tracking algorithm and camera calibration. According to obtained internal and external camera parameters, image pixel coordinates of moving targets and relative position between UAV and moving targets, the UAV world coordinates can be calculated. Multiple sets of experimental data verifies that the absolute error of calculating data is no more than 1.5% of the real data and the relative error is below 2%. Simulation results validate the feasibility of the proposed method in the field of UAV autonomous flight.