Star image recognition based on mutual star distribution

A novel method is designed to divide the star image into polygons, in order to solve the problem of setting the grid size in the grid algorithm. Via the method, each star is located in a corresponding polygon. The geometric features of these polygons are extracted and applied for matching the feature patterns of the guidance star database and the observed star image. Simulations indicate that the novel method removes the possible phenomenon in the grid algorithm where more than one star are located in the same grid. The grid algorithm's disadvantage of recognition failure resulting from identifying incorrect close neighbor star for reference star can be overcome and the accuracy of star image recognition is greatly improved.