Pose estimation of non-cooperative target coated with MLI

Abstract

The relative pose estimation between the chaser and the non-cooperative target is a significant prerequisite for performing on-orbit servicing (OOS) missions. The chaser could design the close-range rendezvous trajectory to approach the non-cooperative target only when the relative pose parameters are obtained effectively and efficiently. In this paper, a pose estimation scheme is designed to obtain the relative pose parameters between the chaser and the non-cooperative target coated with multilayer insulation material (MLI). The scheme utilizes a time-of-flight (TOF) camera to acquire 3D point clouds of the non-cooperative target, and uses the iterative closest point (ICP) method to achieve the point cloud registration between every two frames. Aiming at decreasing the non-systematic errors caused by MLI in the data acquisition of the TOF camera, the corresponding point median filtering approach is adopted to filter out the bad corresponding point pairs generated in the ICP method. A semi-physical experiment is carried out to evaluate the performance of the proposed scheme, the result of which shows that the proposed scheme could notably improve the accuracy of pose estimation of the non-cooperative target, meanwhile the computational efficiency is also ensured.