Improving star centroiding accuracy in stray light base on background estimation

Star sensor, an important component of the spacecraft attitude control system, has the highest attitude measurement accuracy. But the star sensor is often interfered by stray light in space, which affects the star extraction process, resulting in accuracy degradation of the star sensor. To solve this problem, a star extraction algorithm base on background estimation is proposed. By estimating removing the background of the star image, the algorithm simultaneously achieves high star extraction rate and high extraction accuracy. Experiments on both simulated star images and real night-sky images show that the proposed algorithm outperforms state of the art algorithm in both extraction rate and accuracy, especially for the images of very complex background. In addition, computational complexity and fast implementation of the algorithm is also discussed.