Research On Navigation Algorithm Based On Binocular Vision For Soft Landing Asteroid

Because of the communication delay induced by the large distances between the earth and the targeted space explorer, the method of autonomous optical navigation becomes one of the most important navigation ways for landing on an asteroid. The navigation system based on binocular vision technique does not depend on the measure distance from the sensor onboard, and is applied widely in the field of deep space exploration. In this paper, binocular vision based autonomous navigation algorithm are studied deeply for the asteroid landing phase.Firstly, the system of the navigation are analyzed and the coordinates are defined. With projection relation of perspective projection and epipolar geometry considered, binocular vision projection model is built.Secondly, this paper presents an algorithm for feature point detection and matching. Presents a fast SUSAN corner detecting algorithm, the fact that SUSAN edge and corner enhancement uses no image derivatives; a stereo matching approach based on edge feature is proposed, The procedure of the correspondence is fulfilled through the edge-detection, the similarity check and the consistence check, completing the feature tracking and matching for binocular-image.Thirdly, the binocular vision-based motion estimation method is studied. Present a binocular vision-based motion estimation algorithm, this algorithm using binocular measurement method and points-corresponding motion estimation, is an estimate of explorer motion. The output of the algorithm is an estimate of rigid motion between frames.Finally, the structure of asteroid lading navigation system based binocular vision is designed and numerical simulation is performed. The simulation results suggest the navigation system proposed in this paper is feasible, valid and robust.