Optical flow based 3D motion estimation for autonomous landing an UAV on deck

3-D motion estimation method based on computer vision theory is employed to implement a vision guide algorithm for UAV in this paper. First, the image sequences of landing target are taken by the camera mounted on UAV with known focal length and the Lucas-Kanade method is adopted to estimate two successive frame optical flow; then a hierarchical approach is described to effectively decompose the nonlinearities of the 3-D motion estimation into two linear subsystems; finally 3-D motion and structure(depth) information of landing target relative to UAV is recovered without using features of landing target. Experiments using both computer simulated images and real video images demonstrate the correctness and effectiveness of our method.