An Imaging Technique for Safe Spacecraft Landing and Autonomous Hazard Avoidance

Abstract

The focus of this paper is to present an image transformation algorithm to interpret historical imagery land data for autonomous safe landing of a robotic spacecraft. Inherent geographical hazards of Martian terrain may impede safe landing for a science exploration spacecraft. Surface visualization software for hazard detection and avoidance may therefore be applied to enable an autonomous and intelligent spacecraft descent upon entering the planetary atmosphere. The methodology proposed involves integrating linear algebra and computer vision/graphics techniques, with the intrinsic parameters governing spacecraft dynamic motion and camera calibration, in order to assess hazards that might impede spacecraft landing. In this paper, we provide algorithmic details and simulation results of our methodology applied to a representative Mars landing descent profile.