高精度星敏感器星点光斑质心算法

Centroiding algorithm for high-accuracy star tracker

高精度星敏感器星点光斑的质心精度是星敏感器整体精度的基础,它需要达到角秒级的量级。因此,提出了一种基于亚像元相关法的星像质心算法。这种算法是利用互相关匹配实现图像定位的原理,将星像质心定位到亚像素上来提高质心精度。它可以克服由于系统误差和图像采集所带来的误差。根据系统的光学参数设计以及星像的光谱、亮度及其在视场中的位置特征,仿真得到点列图模型,针对视场内与光轴的不同夹角仿真制作一系列亚像元理想模板,然后对星像加高斯随机噪声,把有随机噪声的星像与理想星像进行相关运算来求星像质心精度。通过仿真实验,得到相关法具有较高的星像质心精度,定位精度可以优于1/12个像素的量级。

The centroid accuracy of star tracker is the base of other accuracies. Its pointing accuracy is required to reach arc second scale. A method based on sub-pixel correlation algorithm for star center is proposed. With correlation matching method, the star center is located to sub-pixel level to raise centroid accuracy. The algorithm can overcome system error and error produced by image acquisition. The PSF models are obtained according to optics parameter of system, spectrum of star, luminosity of star and position of star in FOV. In simulation experiment, a series of sub-pixel ideal models is produced corresponding to different incidence angles in FOV. And the Gaussian random noise is added to star spot of different incidence angles. Finally, the accuracy of star center is obtained through the correlation operation for ideal star spot and star spot with random noise. The simulation results show that the accuracy of correlation algorithm is higher than that of centroid algorithm, and it can reach 1/12 pixel scale.