A Fuzzy Logic Approach for Spacecraft Landing Site Selection

Abstract

A method for object identification using a cascading set of two fuzzy inference systems was developed as part of an algorithm to select a landing location for a spacecraft approaching the surface of the Moon. The method relies only on optical images, eliminating the need for lidar, and utilizes the tolerance for uncertainty inherent to fuzzy systems to reduce the amount of data that must be processed, thus reducing the computational burden. The algorithm is tested on a dataset of 36 real lunar images in which a human analyst has created labels for each large rock seen in the image. The algorithm was able to identify 80% of the labeled rocks. The algorithm also classified additional features as rocks, shadows, or sky that were not selected by the human analyst. Using this broader definition of hazards, 72% of the cases found an acceptable landing site.