BRISK based target localization for fixed-wing UAV's vision-based autonomous landing

摘要

In this paper, a Binary Robust Invariant Scalable Keypoints (BRISK) based detection is utilized to facilitate the flying unmanned aerial vehicle (UAV) localization within its autonomous landing on the runway. Specifically, two target detection algorithms are proposed and developed as the BRISK-supported approach. Dataset of images and differential GPS are recorded by a ground stereo vision guidance system. By virtue of the landing dataset, experiments are conducted then. Experimental results validate the effectiveness, in terms of efficiency and accuracy, of the proposed target localization algorithms for the fixed-wing aerial vehicle's autonomous landing.