星三角形识别中的选星策略及实现

Star selection strategy and realization in star triplet identification

构造观测星三角形时,若其中某星像不能与星表存储的导航星对应,则该三角形无法正确匹配。为避免该星像被用于构造其它星三角形影响识别速度,讨论了此时星三角形的选星策略问题。提出了一种同时考虑选星子集相似性与子集中星像亮度两个条件的选星策略,给出了基于相似性排序和权值渐消的实现算法,得到了相邻星三角形选星子集差别最大的结果子集序列。仿真比较了6个星像(内含4颗导航星)时,采用该序列选星与采用参考序列选星,完成正确星图识别所需的最大选择次数,结果表明两者最大相差6,平均相差1。设定不同星像个数和导航星个数时两序列相差结果,平均约为1。说明该子集序列对提高星图识别速度是有价值的。

If a certain star in the star triplet has no corresponding guide star in star catalog,the star triplet can not be correctly identified.To prevent this star be used to construct next star triplets,the problem of star selection strategy is discussed.A selection strategy based on subset comparability and star brightness in star selection subset was proposed.Its realization algorithm based on comparability ranking and weight decreasing was given,and the resulted series with maximum difference in selected star subset of adjacent star triplet were obtained.In simulation,the maximum running times to get the identifiable star triplet in the image with 6 stars（4 guide stars inside） by the resulted series were compared with those by the reference series.The results showed that maximum difference was 6 and average difference was 1.When the stars and the including guide stars were set to different numbers,the difference between the two series was about 1 on average.The result is valuable for improving the speed of star triplet identification.