Iterative algorithm for autonomous star identification

An autonomous star identification algorithm is described in this study. This algorithm iteratively determines a catalog star, the distances of which to its neighbors are similar to those of a given sensor star and its neighboring stars, until a unique match is found. In each iteration, only the catalog stars that receive sufficient votes are used in the subsequent steps. We also develop a useful probabilistic model for performance prediction and evaluation. Analysis indicates that spurious star pairings are decreased exponentially after each iteration. The simulation and real sky image results show that the proposed approach is more robust than conventional algorithms and requires fewer resources.