Full-sky autonomous star identification based on radial and cyclic features of star pattern

An algorithm for full-sky autonomous star identification is presented in this paper. Based on the radial and cyclic features, a star pattern is successfully generated, and a guide database is also established according to the generating process of the star pattern. Since the radial pattern is a kind of reliable rotation-invariant feature independent of other unreliable information, it is used for the initial match. And then the cyclic feature is adopted for the follow-up match. Meanwhile, some constraints are introduced to eliminate the redundant and wrong star pairings. In simulations, using the stars brighter than magnitude 6.0 Mv and a 12 × 12 deg field of view (FOV), the algorithm obtains an identification rate of 97.57% from the statistics of 10,000 random sensor orientations at a positional noise level of 1 pixel.