In order to support spacecraft proximity operations, such as on-orbit servicing and spacecraft formation flying, several vision-based techniques exist to determine the relative pose of an uncooperative orbiting object with respect to the spacecraft. Depending on whether the object is known or unknown, a shape model of the orbiting target object may have to be constructed autonomously by making use of only optical measurements. In this paper, we investigate two vision-based approaches for pose estimation of uncooperative orbiting targets: one that is general and versatile such that it does not require a priori knowledge of any information of the target, and the other one that requires knowledge of the target's shape geometry. The former uses an estimation algorithm of translational and rotational dynamics to sequentially perform simultaneous pose determination and 3D shape reconstruction of the unknown target, while the latter relies on a known 3D model of the target's geometry to provide a point-by-point pose solution. The architecture and implementation of both methods are presented and their achievable performance is evaluated through numerical simulations. In addition, a computer vision processing strategy for feature detection and matching and the Structure from Motion (SfM) algorithm for on-board 3D reconstruction are also discussed and validated by using a dataset of images that are synthetically generated according to a chaser/target relative motion in Geosynchronous Orbit (GEO).
展开▼
