An Accelerated Motion Blurred Star Restoration Based on Single Image

Abstract

The accuracy of star centroid extraction and identified star number is both crucial features for star sensor precision. Motion blur and fracture are introduced when star sensor works under high dynamic conditions, which affect the accuracy of star centroid extraction and further reduce the precision of the star sensor. To improve the precision of star sensor, this paper proposes a star image restoration algorithm, including blur kernel estimation as well as an accelerated Richardson–Lucy (RL) reconstruction for motion blur star image under high dynamic conditions. First, an improved Radon transform method is presented by introducing a combination of Z-function and double threshold mask that have considerable anti-noise performance. Based on this improved method, the blur kernel of the degraded star image can be obtained by only utilizing a single blurred star image. Furthermore, as the traditional RL is time-consuming, to overcome this shortcoming, an accelerated algorithm based on the second-order vector extrapolation is proposed, offering speedup of 20 times over the original RL and five times over the existing acceleration algorithms. Finally, experiments on simulated as well as real star images demonstrate that the proposed algorithm is effective in improving the dynamic performance of star sensor even under low signal-to-noise ratio conditions, which is of great importance for the further applications of star sensor.