Research on Star Point Cluster Location Method

Abstract

Various location methods are proposed for the star point cluster images of the interferometric star sensor. They include global-centroid method, circle fitting method and star point average method. The global-centroid method takes the star point cluster as a whole for grayscale weighting calculation. The circle fitting method locates the center of the circle obtained by secondary star points circle fitting. The star point average method averages the coordinates of total sub-star points to realize location. Under the simulation conditions in this paper, in the noiseless star map, the location errors of all kinds of star point cluster location methods are lower than that of the single star point centroid method, and the global-centroid method is three orders of magnitude less than that of the single star point centroid method. In the star map with Gaussian noise, the location error of the star point average method is the smallest, and it is 34% less than that of the single star point centroid method.