Stereovision-based Noncooperative Spacecraft Pose Measurement via Circle and Planar Points

In this paper, a novel approach based on the feature circle and planar points is proposed for the relative pose measurement of the noncooperative spacecraft. Specifically, the approach uses information from the stereovision-based navigation system to provide multiple perspective projections of the circle and planar points on the noncooperative spacecraft. Then the circle center position is recovered and the normal direction is determined by duality disambiguation. Furthermore, a point depth of recovery from the stereovision is used to define the direction from the circle center to the point, which is regarded as the initial attitude for the stereovision-based orthogonal iteration pose estimation to avoid the local minimal value of the iteration method. Therefore, by utilizing the initial attitude and the orthogonal iteration algorithm based on planar points, the pose ambiguity from the planar points is eliminated and the roll angle around the circle normal direction is available. Finally, numerical simulations are conducted to validate the effectiveness of the proposed algorithm.