一种基于旋转不变特征的星图识别算法  

A Novel Star Pattern Identification Algorithm Based on Rotation-invariant

作　　者：胡坤[1] 陈嘉鸿 谢元平[1] 

机构地区：[1]国防科学技术大学光电科学与工程学院,长沙410073 [2]飞行器海上测量与控制联合实验室,江苏江阴214431

出　　处：《光电工程》2015年第6期20-26,共7页Opto-Electronic Engineering

基　　金："飞行器海上测量与控制技术"联合实验室开放基金(FOM20140F012)

摘　　要：栅格算法作为星图识别算法中的代表,有着识别速度快、识别率高、导航数据库容量小的优点,应用较广,但其在构造恒星特征模式时存在需要选取临近星进行图像旋转的问题。为克服这一缺陷,提出了一种基于旋转不变特征的星图识别算法。该算法利用模式识别类星图识别算法思想,为每颗恒星建立一种与旋转无关的特征模式,采用改进的Hausdorff距离进行匹配识别。仿真实验结果表明,基于旋转不变特征的星图识别算法对星点位置噪声与缺失星鲁棒性均优于栅格算法,具有良好的识别效果,但其识别速度与存储容量仍需进一步优化。As a typical algorithm for star identification, grid algorithm has advantage of fast identification time, high identification rate and small capacity of guide star database, but it has the problem of rotating the picture with a chosen neighbor-star when creating the star pattern. In order to solve the problem, a star identification algorithm based on rotation-invariant is proposed. The algorithm uses the star identification algorithm based on idea of pattern-recognition to build a rotation independence pattern for every star, and uses the improved Hausdorff distance for the matching. The simulation results show that identification algorithm based on the rotation-invariant based star has a better robustness in position errors and lost stars than grid algorithm, and the algorithm has good recognition effect, but the identification time and memory capacity still need to be optimized.

关 键 词：星图识别 栅格算法 HAUSDORFF距离 

分 类 号：V448.22[航空宇航科学与技术—飞行器设计] TP391.41[自动化与计算机技术—计算机应用技术]