Autonome monokulare optische Identifikation von Mondkratern aus unbekannter Kameraposition und -Lage

Abstract

Present and future robotic landing missions to the moon with high landing accuracy requirements need a capability of autonomous position and attitude determination. Optical sensors provide a cheap and robust means of generating image measurements, and lunar craters provide persistent landmarks. This dissertation describes a method that allows inferring a full camera pose from a single image of the lunar surface under a wide variety of illumination conditions and for a previously unkown position and attitude of the camera.