抄録
Obstacle detection and avoidance are essential for achieving a successful soft landing on the moon. In a previous paper, we proposed using local and global statistics of image intensities to select a safe landing site. In this paper, we apply that method to photographs of the moon's georama. In the assumed landing scenario, an unmanned lander takes photographs of the landing area, detects obstacles, and makes avoidance maneuvers. In the experiments, our image processing method is applied to the photographs of the georama taken at three different distances, and the validity of the finally selected site is evaluated. It is shown that safe landing sites are selected in conditions with a variety of light directions, and the robustness of the proposed method is confirmed.
