天顶筒测量世界时的CCD星图处理方法

CCD star map processing method for measuring UT1 with zenith telescope

为实现UT1自主测量,中国科学院国家授时中心建立了基于新型数字照相天顶筒的UT1测量网络。该新型天顶筒采用CCD(charge coupled device)代替传统胶片作为感光元件,以提高观测效率和测量精度。本文分析了天顶筒CCD星图噪声特性,发现以高斯噪声为主,对比了多种滤波算法,发现高斯滤波算法对星图的降噪效果更好。根据实拍星图灰度分布的星像区域远小于背景噪声区域、星像灰度值高于背景噪声灰度值等特征,选取合适阈值对整幅星图进行分割,很好地将星像从背景噪声中分离出来。比较了多种星像质心位置的计算方法,发现加权质心算法效果更好。通过实测数据检验,建立了一套适用于天顶筒星图处理的方案。

In order to realize the autonomous measurement of UT1,the National Time Service Center of Chinese Academy of Sciences established a Digital Photographic Zenith Telescope network.In order to improve the observing efficiency and measurement accuracy,the Digital Photographic Zenith Telescope uses CCD instead of the traditional film.In this paper,we analyzed the noise characteristics of the CCD star map,and found that Gaussian noise is the main noise source.So we compared a variety of filtering algorithms,and among them the Gaussian filtering algorithm exhibited better noise reduction performance than the others.According to the gray distribution characteristics of the real star map,the star image region is far smaller than the background noise region,and the star image gray value is higher than that of the background noise,we can select appropriate threshold value to segment the whole star map,which could separate the star image from the background noise well.In this paper,we compared several methods for calculating the centroid position of star image,and the weighted centroid algorithm was found to be more effective.Based on the research,we developed a set of optimal scheme for star map processing.