基于星三角形外接圆的星图识别算法  

Star pattern identification algorithm based on star triangle circumcircle

作　　者：文涛[1] 王宏力[1] 陆敬辉[1] 詹巧林[1] 高志勇[2] 

机构地区：[1]第二炮兵工程学院304教研室,陕西西安710025 [2]空军驻华北地区军事代表室,北京100086

出　　处：《传感器与微系统》2012年第2期74-76,80,共4页Transducer and Microsystem Technologies

摘　　要：三角形外接圆半径能部分描述三角形的结构特征,在三角形两条边长及其外接圆半径已知的情况下,可以确定唯一的三角形,因此提出了一种基于三角形外接圆的星图识别算法。构建了导航星数据库,以特征三角形为匹配模板,减少了导航星三角形的数量,从而减小导航星数据库容量。为了提高搜索效率,以特征半径为搜索量,并对其进行升序排列。通过对特征半径的匹配,缩小了角距匹配的范围,提高了角距匹配的速度,同时采用的多三角形的组合有效地提高了识别率。为了保证星图识别的准确性,引入了验证识别环节。仿真结果表明:当存在2像元的位置噪声时,识别率大于97.42%,平均识别时间为38.41 ms,实时性与鲁棒性均优于传统三角形星图识别算法。A new star pattern identification algorithm is proposed based on the fact that any triangle is determined by its two sides and circumcircle radius which can partially describe the structural feature of the triangle. Navigation star database is constructed using this algorithm, the capacity of the navigation star database is reduced by using characteristic triangles as the matching template. The characteristic raduis is used for indexing and carried on a ascending rank to improve the searching efficiency in the process of star pattern identification, the matching scope of the argument is narrowed and the matching rate of argument is improved by the matching of the characteristics raduis, and the identification rate is effectively improved by the combination of multi-triangles. The verification process is introduced to confirm the identification results. Simulation results show that the proposed algorithm has better robustness for position error, when the Gauss noise of position error is two pixels, the identification rate is larger than 97.42 % and the average identification time is 31.41 ms.

关 键 词：三角形外接圆 星图识别 导航星数据库 特征三角形 特征半径 

分 类 号：U324.24[交通运输工程]