Performance evaluation of a vision sensor in 3D virtual environment for rendezvous and docking application

摘要

In the international scenario of nations developing space stations, rendezvous and docking [RVD] is one of the most important processes associated with astronavigational activities. During the process of docking, as the distance between chaser and target spacecraft decreases on the approach line to the target spacecraft, the lateral displacement and velocities relative to the target will also decrease proportionally. This necessitates increased navigational accuracies along the approach sequence towards achieving docking. The absolute/relative navigation sensors used have a limited range and accuracy of operation, which determines the extent of the particular approach phase in which it will be used. Vision sensors can provide millimetre accuracies in less than ten meter range. This made Vision Sensors/ Computer Vision/ Machine vision a key element in proximity phase of RVD missions, especially in automated rendezvous and docking. This paper proposes a vision sensor for the above application and validates the concept and accuracy requirements using a 3D virtual environment.