A two-step matching algorithm for autonomous star identification

Abstract

Abstract. An algorithm based on the primary star, called the two-step matching algorithm, is proposed. The recognition characteristic of this algorithm consists of angular distances between the primary star and its neighboring stars, and the directions of neighbor stars. This algorithm uses star map information with high efficiency. The performan ce of the algorithm is analyzed through a simulation experiment that compares it with grid algorithm and another algorithm based on the primary star. Simulation results show that the correct identification rate of the two-step matching algorithm is more than 98% when the standard deviation of Gaussian noise is 2 pixels. This algorithm is robust with fake stars and magnitude deviation. The identification rate is more than 96% when five fake stars are present in the star maps. In addition, the algorithm has many advantages, such as quick identification, being unaffected by magnitude noise, and having a small guide star catalog. The two-step matching algorithm has good application value and is promising in the area of star identification.