Attitude information deduction based on single frame of blurred star image

The paper presented a research on the star image´s processing acquired by star sensors in a high-dynamic situation; considering that the star image gained by star sensor would be dramatically blurred when in high dynamic situation, fast rotating state conventionally, and could not be applied directly in succeeding process, the author proposed a procedure to deduce the star tracker´s attitude information with a single frame of blurred star image. Based on the Gaussian PSF model of the star, the paper proposed a routine to detect the blur length in the region of interest (ROI), and therefore the rotating velocity and other attitude information could be figured out with a Least Square filtering. With plenty simulation and experiments, the paper proved some certain practicability of the algorithm.