OPTICAL TERRAIN NAVIGATION FOR PINPOINT LANDING: IMAGE SCALE AND POSITION-GUIDED LANDMARK MATCHING

摘要

Map-based accurate navigation is a requirement for future planetary missions requiring a pinpoint landing capability. Based on an Extended Kalman Filter tight fusion architecture between a camera and inertial sensors, this paper proposes a new method to match descent image landmarks with orbital ones down to very low altitude while being robust to 3D terrain topography and illumination changes. It is based on the idea that image scale can be used as an additional description for a landmark. Filter estimates are used to predict the research area and scale of the landmarks in the descent image. A Monte Carlo analysis using software simulation shows its compliance with pinpoint landing requirements.