Motion parameters estimation algorithm of star sensor based on Zernike moments under dynamic conditions

Abstract The dynamic performance of star sensors is mainly affected by smearing effect of star spots. In order to improve the dynamic performance, it is necessary to perform blur kernel estimation and restoration of star image. However, the accuracy of motion parameters estimation of existing algorithms is not high enough, which limits the effect of restoration. An efficient motion-blurred star image's motion parameters estimation algorithm based on Zernike moments is presented to promote the accuracy of centriod extraction and attitude estimation under dynamic conditions. The proposed algorithm improve the dynamic performance of star sensors through two aspects. First, a unitized roof model is established for the dynamic star spots. Second, the Zernike moments of unitized roof model is derived, and the sub-pixel edge detection method is used to estimate the angle and length of star streak to obtain motion parameters of the star image. The effectiveness and excellent performance of proposed algorithm is verified by simulations and ground experiment. Experiments show that this algorithm is robust to noise, improves the estimation accuracy of blur parameters, and thus improves the accuracy of star spot extraction. Meanwhile, the algorithm has less time consumption, and is suitable for improving the dynamic performance of star sensors.