星点图像的目标提取算法研究

Research on Star Area Extraction Algorithm from Star Images

摘要:

对于星点灰度重心法等亚像素提取算法来说,需要提取星点区域内的每一个像素。基于扫描的边界搜索法只能获取目标的边界像素,文中对其进行了扩展,通过判断边界像素中最大最小行列数来确定目标区域的具体范围,从而提取目标区域内的每一个像素。针对大视场星点图像目标范围小、相互间距离远的特点,为减小计算量,提出了基于视框搜索的目标提取算法。通过实验证明,基于视框的算法在计算量上比基于边界搜索的算法减少3/4。

Abstract:

Star area extraction is necessary before calculating the sub-pixel center of star in the star image processing of star sensor. It is necessary to get every pixel in the star area in star image for some sub-pixel algorithm. In this paper, the boundary searching method that can only get the boundary pixels was developed to get every pixel in star area according to the max and min line number and column number. To reduce the computing amount, the search method based on visual case was represented for the wide FOV star images. The calculate amount of visual case searching method was reduced by 3/4 compared to the boundary searching method according to experiments.