CHEN Hao, MA Yu-kun, FENG Hui. A new image feature recognition method based on space vehicle end surface feature[J]. 2016, 35(9): 46-48+53. DOI: 10.13873/J.1000-9787(2016)09-0046-03.
航天运载器端面特征的新型图像特征识别方法
摘要
基于航天运载器端面的固有特征
结合现有航天图像测量系统的硬件平台
提出一种端面图像特征提取与识别方法
获取视场内关键目标端面特征信息
对后续图像物理量参数优化解算起到指导性作用。以运载器端面椭圆特征为基础
利用Canny算子检测出图像中端面椭圆边缘特征
将特征边缘拆分为若干弧段
通过弧段拟合、弧段合并、椭圆提取等方法
实现端面特征的准确提取。仿真试验结果表明:该图像特征识别方法实时性好、精度高
可显著提高运载器图像测量能力。
Abstract
Based on the inherent characteristics of space vehicle end surface
combined with the existing space image measurement system hardware platform
propose a method for end surface image feature extraction and recognition
to obtain the key objectives of surface feature information of the field and the follow-up image physical parameters optimization solution plays a guide role.On the basis of space vehicle elliptic edge feature
the ellipse edge feature is detected using the Canny operator
then the edge feature is split into several arcs.Using the method of arc fitting
arc merger and ellipse extraction
end sur face feature is extracted accurately.Simulation results show that the image feature recognition method has good real-time performance and high precision
and can significantly enhance image measurement ability of space vehicle.