Considering that the characteristic pattern extracted by traditional grid algorithm can ’t reflect its internal similarity degree

which can affect the accuracy of star map recognition

a modified grid algorithm is proposed. The algorithm uses the metric functions derived from the characteristic patterns to reflect the similarity degree between different characteristic patterns

which solves the problem exists in feature mode extraction because influence of the position measurement error. Simulation environment according to the real situation is set up. The results indicate the star map identification success rate of the proposed method is obviously higher than the traditional algorithm when the star point position error is greater than 1. 5 pixels and the improved algorithm of star map recognition success rate is superior to the existing algorithms in the presence of"artificial star"

which shows the validation of the proposed algorithm.