极值分析下基于位置矩阵的质心定位方法  

Centroid location method based on position matrix under extremum value analysis

作　　者：段辉 张志利 周召发 梁哲 姚鑫玲 DUAN Hui;ZHANG Zhili;ZHOU Zhaofa;LIANG Zhe;YAO Xinling(School of Missile Engineering,Rocket Force University of Engineering,Xi’an 710025,China;State Key Discipline Laboratory of Armament Launch Theory and Technology,Rocket Force University of Engineering,Xi’an 710025,China)

机构地区：[1]火箭军工程大学导弹工程学院,陕西西安710025 [2]火箭军工程大学兵器发射理论和技术国家重点学科实验室,陕西西安710025

出　　处：《系统工程与电子技术》2021年第9期2594-2604,共11页Systems Engineering and Electronics

基　　金：国家自然科学基金(41174162)资助课题。

摘　　要：为以使星敏感器在复杂工况下依旧能获得高精度的星点质心位置,提出一种极值分析下基于位置矩阵的质心定位方法。首先,提出极值分析下的星点弥散区域粗提取算法,对去噪后的星图进行星点局部区域限定。接着,提出基于位置矩阵的线段编码标记法对星点弥散区域进行精提取。最后,利用星点区域像素信息估计出高斯曲面参数以得到星点精确的质心坐标。仿真结果表明,星图噪声标准差处于2的水平时,所提方法可以使星点质心定位精度达到0.0704 pixel,且在星图噪声标准差处于12的水平时,仍具有90.03%的高星点提取成功率。In order to obtain the high-precision star centroid position under complex working conditions,a centroid location method based on position matrix under extremum value analysis is proposed.Firstly,a coarse extraction algorithm of star point dispersion region is proposed under extreme value analysis,and the star point local region is restricted after denoising.Then,a line segment coding labeling method based on position matrix is proposed to extract the star dispersion region accurately.Finally,Gaussian surface parameters are estimated by using star pixel information to obtain the precise centroid coordinates of star.Simulation results show that when the star map noise standard deviation is 2,the accuracy of the centroid location can reach 0.0704 pixel by the proposed method,and when the star map noise standard deviation is 12,the success rate of star extraction is 90.03%.

关 键 词：极值分析 位置矩阵 线段编码 质心定位 

分 类 号：V448.22[航空宇航科学与技术—飞行器设计]