光学 精密工程, 2017, 25 (6): 1627, 网络出版: 2017-07-10   

添加补偿码的快速径向伴星特征星图识别

Radial neighbor feature with compensate code star pattern recognition algorithm
作者单位
中国科学院长春光学精密机械与物理研究所, 吉林 长春 130033
摘要
针对传统的基于径向特征的星图识别算法在构建星模式的过程中由于位置噪声的干扰导致识别率较低的问题, 本文提出一种添加补偿码的快速径向伴星星图识别算法。该算法以比特向量的形式构建基于径向特征的特征向量, 同时将伴星间的角距信息以及位置噪声的补偿信息添加到特征向量中, 从而有效地减小了特征库的容量, 提高了星图识别算法的稳定性和识别率。最后本文根据比特向量的特点采用最小相似差方法快速完成观测星与导航星之间的初匹配, 再根据同一视场内星点位置信息的相关性完成对观测星的唯一识别。实验仿真结果表明, 在位置噪声为0.5像素的情况下星图识别成功率达到97.8%; 在星等噪声为0.8 Mv的情况下星图识别成功率达到96.4%; 当以真实星图为实验对象时, 星图识别的成功率达到94.2%。与传统的三角形算法以及未添加补偿码的径向特征星图识别算法相比, 本文算法在识别成功率和识别时间上均有着不同程度的提高。
Abstract
A fast radial featured star pattern recognition algorithm with compensate code was proposed to solve star model that constructed by traditional radial featured star pattern recognition algorithm could be vulnerable to position noise, lead to low recognition rate. This algorithm adopt byte vector form to construct radial feature vector, meanwhile neighbor angle distance and compensate information were added to feature vector to reduce guide star catalogue capacity and to effectively increase stability and recognition rate of star pattern recognition algorithm. In addition a minimum similar difference matching method was formulated according to the characteristics of byte vector thus completing initial matching between the observation star and the guide star. Fathermore,the characteristic of position information coherency of stars in the same field of view was applied for accomplish exclusive recognition. The result of the experiment shows that recognition rate is 97.8% under position noise of 0.5 pixel ,and under magnitude noise of 0.8 Mv is 96.4%. For recognition of real star images, the recognition rate of this algorithm achieves 94.2%. Compared with the triangle algorithm and the traditional radius recognition algorithm,the recognition rate and time of this algorithm are improved in various degrees.

高扬, 赵金宇, 陈涛, 王敏. 添加补偿码的快速径向伴星特征星图识别[J]. 光学 精密工程, 2017, 25(6): 1627. GAO Yang, ZHAO Jin-yu, CHEN Tao, WANG Min. Radial neighbor feature with compensate code star pattern recognition algorithm[J]. Optics and Precision Engineering, 2017, 25(6): 1627.

本文已被 1 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!