Star image processing method of star sensor based on star points prediction

Star image processing method based on star point prediction is proposed, to solve the problems of large storage space of star image data, large computation of star point extraction and fake stars that cannot be used for attitude determination in traditional star sensor in this paper. According to the attitude prior information, the navigation star library is used to predict the position of stars in the detector and the stripe information of stars. Under the constraints of the total number of pixel rows and the total number of stripes, only the stripes with more stars are reserved for the design of the detector driver by clustering merging. At the same time, the scan line seed filling method is used to extract stars only in the predicted star window region, which greatly reduces the image storage space and star image extraction time consumption. The experimental results show that compared with the traditional full image processing method, the star sensor image processing method based on star point prediction has great advantages in star point extraction stability, storage space and time consumption.