基于惯性比特征的三角形星图识别算法  

Triangle star identification algorithm based on inertia ratio

作　　者：樊巧云[1] 陆壮志[1] 魏新国[1] 李健[1] 

机构地区：[1]北京航空航天大学仪器科学与光电工程学院精密光机电一体化技术教育部重点实验室,北京100191

出　　处：《红外与激光工程》2012年第10期2838-2843,共6页Infrared and Laser Engineering

基　　金：教育部新世纪人才支持计划(NCET-10-0037);长江学者和创新团队发展计划(IRT0705)

摘　　要：为了提高星敏感器的星图识别性能,提出了一种基于惯性比特征的三角形星图识别算法。该算法将导航三角形的形心惯性比和导航三角形最长边的角距值作为匹配特征量,并以此构建导航特征库,采用散列函数的方式搜索导航特征库。由于将三角形特征量的维数由三维缩减至二维,该算法具有导航星数据库容量小、搜索匹配速度快的优点。仿真实验表明,在星点位置噪声标准偏差为2像素、星等噪声标准偏差为0.8等的条件下,该算法的识别率均在99%以上,平均识别时间约为1.29ms,和现有三角形算法相比均具有一定的优势。并在地面观星实验中,取得了非常好的效果。To increase the performance of star identification for star sensor, triangle star identification algorithm based on the character of inertia ratio was proposed in this paper. This algorithm extracted two- dimensional features of triangle with maximum angular distance and ratio of triangle centroid inertia, and then built a guidance characteristic catalogue which was searched by means of hash function. Owing to the depression of triangle character from three-dimension to two-dimension, the algorithm possessed the advantages of less capacity of guidance database and faster speed. Simulation shows that this algorithm can reach 99% on identification rate when exists standard deviation of two pixels star position noise or 0.8 Mv star magnitude noise in the star map, and the average identification time is 1.29 ms. And this algorithm has definite superiority compared with existing triangle algorithms. Moreover, there is a good result for the experiment of star observation in the ground.

关 键 词：星敏感器 星图识别 三角形算法 惯性比 

分 类 号：V448.22[航空宇航科学与技术—飞行器设计]