Autonomous rendezvous and docking is necessary for planned space programs such as DARPA ASTRO, NASA MSR, ISS assembly and servicing, and other rendezvous and proximity operations. Estimation of the relative pose between the host platform and a resident space object is a critical ability. We present a model-based pose refinement algorithm, part of a suite of algorithms for vision-based relative pose estimation and tracking. Algorithms were tested in high-fidelity simulation and stereo-vision hardware test bed environments. Testing indicated that in most cases, the model-based pose refinement algorithm can handle initial attitude errors up to about 20 degrees, range errors exceeding 10% of range, and transverse errors up to about 2% of range. Preliminary point tests with real camera sequences of a 1/24 scale Magellan satellite model using a simple fixed-gain tracking filter showed potential tracking performance with mean errors of < 3 degrees and < 2% of range.
Vision-based relative pose estimation for autonomous rendezvous and docking
2006 IEEE Aerospace Conference ; 20 pp.
2006-01-01
1194441 byte
Conference paper
Electronic Resource
English
Relative Pose Estimation under Monocular Vision in Rendezvous and Docking
Trans Tech Publications | 2013
|Gyro-Aided Vision-Based Relative Pose Estimation for Autonomous Rendezvous and Docking (AAS 13-101)
British Library Conference Proceedings | 2013
|Relative Angles-Only Navigation and Pose Estimation for Autonomous Orbital Rendezvous
Online Contents | 2007
|