Fast Matching Algorithm for Sparse Star Points Based on Improved Delaunay Subdivision

Abstract

The image under the starry sky background lacks texture detail information. It is difficult to use visual features such as regional features, shape to achieve image registration, and the star map structure may be unknown. To address these issues, this paper proposes a fast matching algorithm for sparse star points based on the Delaunay subdivision, which uses the geometric topological structure between the star points to solve the image transformation parameters and achieve accurate registration. Experimental results show that this method can still perform correct registration even in the presence of noise, target points, rotation, and translation, or lack of star points in the star map. The average registration error of 50 sets of simulated star maps is 0.4791 pixels, and the average registration time is 3.5386 s, which can meet the requirements of registration accuracy and speed, realizes the combination of mathematical methods and graphics.