Learning Illumination Invariant Features for Lunar South Pole with Deep Learning

A major challenge for vision-based applications of planetary exploration is large illumination variance that is caused by a combination of the sun position, terrain morphology, and lack of light scattering due to the absence of an atmosphere. This diminishes the capability to recognize terrain features or landmarks that are critical for autonomous robotic operations, including spacecraft pinpoint landing and navigation. In this paper we explore deep learning for learning illumination invariant features in the challenging lunar domain.