Pose estimation and trajectory calculation for autonomous close-range rendezvous and docking

Abstract

In order to examine secure ways to perform autonomous satellite servicing applications, this paper provides an approach for close-range motion estimation for rendezvous and docking scenarios. A stereo-camera based method is presented, which generates estimates of the unperturbed motion performed by a completely uncooperative target. No a-priori knowledge about the target has to be provided to accomplish the motion estimation task. A state of the art simulation system is utilized. It provides rigid body dynamics as well as realistic lighting conditions and noise models for the sensor simulation and thus enables the setup of virtual testbeds. Two different satellite models were used to verify the functionality and accuracy of the approach described in this paper. Two KUKA LWRs were used to actuate a model of a Galileo satellite and a stereo-camera rig. This way the method was verified in a physical environment, as well.

Relative Attitude Estimation for a Uniform Motion and Slowly Rotating Noncooperative Spacecraft

“…In other words, the target is absolutely a noncooperative object. Priggemeyer et al [14] acquired the target picture sequence by a stereo camera system with a baseline of 0.2 m. Afterwards, they extracted the point features with SURF. And then the DLT algorithm [15] solving a linear equation system was applied to find the projection matrices.…”