Drogue recognition and location for UAV Autonomous Aerial Refueling based on camera calibration

In order to improve the docking success rate in Autonomous Aerial Refueling(AAR), and reduce the calculation error of drogue position, the camera calibration method has been combined with MeanShift-KalmanFilter (MS-KF) algorithm in this paper to solve the problem of drogue recognition and location. Based on the pin-hole camera model, adaptive corner detection method was applied to pick up feature points in calibration board. Then, three camera calibration external factors had have been investigated by contrastive experiments. The conclusion of the above experiments was utilized to initialize camera calibration conditions. Finally, MS-KF algorithm was introduced in hardware-in-loop test for drogue recognition and location. The experimental results validate the effectiveness of the proposed algorithm.