星点识别算法中背景阈值的确定方法  

The Method of Deciding Background Threshold in Star Dot Recognition Algorithm

作　　者：罗长洲[1] 王胜男[1] 邵雷[1] 赵广州[1] 王杰[1] 

机构地区：[1]北京控制与电子技术研究所,北京100038

出　　处：《航天控制》2013年第5期50-54,共5页Aerospace Control

摘　　要：数字星空图像是星敏感器加工处理的对象,其特点与目标和目标所在星空区域、图像传感器的光谱响应特性,以及捕获图像时的工作条件等有关。选择合适的背景阈值,是正确识别星点的基础。在分析总结数字星空图像基本特点的基础上,根据数理统计和误差分析理论,提出一种背景阈值确定方法,又根据星敏感器光电探测过程中的物理模型,提出因果关系法确定背景阈值的技术路线。依据特定方案星敏感器的试验结果表明,采用因果关系法确定的背景阈值,能很好地将数字星空图像中的背景和星点区分开来,确保了系统对目标星点的正确识别,并且具有更好的实时性。The digital star image is the object of star sensor processing, and its characteristics are decided by the target star, the area around the target star, the spectral response of image sensor and the work condi- tion of the special star sensor. The appropriate background threshold is the basis to recognize the star dots correctly. According to the theory of statistics and error analysis, the characteristic of digital star image is analyzed and a new method on background threshold is proposed, which is based on causality. The experi- mentation is done to the special star sensor, and the result shows that the star dots have been differed from the background in a digital star image. This ensures that the star sensor can recognize the star dots correctly and can work in real time.

关 键 词：星敏感器 数字星空图像 点目标 识别 阈值 

分 类 号：TP752.1[自动化与计算机技术—检测技术与自动化装置]