Comparison Of Fast Catalog-Stellar Image Matching Algorithms

The position and brightness of stars are important information to the astronomical research and space navigation. To facilitate data calculation and processing, We usually need to match fast the information (such as position and brightness et al.) extracted from CCD images of a dense star field to the corresponding information from a stellar catalogue. Therefore, a variety of catalogue matching algorithms have been developed. Otherwise, in space navigation, using CCD camera to get the image and extract star position and brightness et al from fields of view, then using matching algorithm to find the corresponding matched stars of the observed stars from the guide catalogue, finally with these matched stars to calculate the position of the spacecraft. In this paper, we compare three latest matching algorithms:i) method based on the triangle, ii) method based on the vector; iii) method based on the radial and cyclic features of star pattern.Through matching the CCD images for open clusters NGC2168 and NGC1664 taken at the Yunnan Observatory by the 1 m telescope, we compared the construction and matching phase of the whole matching progress, and the result shows that the method based on the triangle costs a lot of time at the triangle construction, the method based on the radial and cyclic features of star pattern is slow at matching phase, and only the method based on the vector cost least time at these two stages. In addition, we also compared the matching rate for different size of lists. The method based on the triangle is the highest, the method based on the vector is middle, its matching rate reach 100% when a list size is more than 30, the matching rate of the method based on the radial and cyclic features of star pattern is lower than the others. Therefore, the method based on the vector exhibits faster matching speed and it thus is better.