首页 | 官方网站   微博 | 高级检索  
     

基于不变特征的标志识别技术研究
引用本文:魏祥泉,李金宗,李冬冬.基于不变特征的标志识别技术研究[J].光电工程,2007,34(1):13-18.
作者姓名:魏祥泉  李金宗  李冬冬
作者单位:哈尔滨工业大学,航天学院,黑龙江,哈尔滨,150001
摘    要:标志识别是近距离空间交会对接中的一项重要的关键技术.在空间交会对接过程中,太阳和月球等各种天体以及飞行器上零部件的辐射与反射都会影响CCD的成像,形成大量杂散光干扰.为此,本文首先设计了具有不变特征的标志系统,然后根据标志系统固有的不变特征,详细设计了左右像点匹配、远场像平面不变特征识别、近场像平面预测识别以及像空间模型匹配全等识别等一套完整算法,可以保证彻底消除杂散光的干扰,正确可靠地完成各个标志灯像点的辨识.模拟图像仿真实验和真实图像仿真实验结果验证了算法的正确性和可靠性.

关 键 词:标志识别  不变特征  像点匹配  预测识别
文章编号:1003-501X(2007)01-0013-06
收稿时间:2006/4/11
修稿时间:2006-04-11

Beacon recognition technique based on invariant features
WEI Xiang-quan,LI Jin-zong,LI Dong-dong.Beacon recognition technique based on invariant features[J].Opto-Electronic Engineering,2007,34(1):13-18.
Authors:WEI Xiang-quan  LI Jin-zong  LI Dong-dong
Affiliation:School of Astronautics, Harbin Institute of Technology, Harbin 150001, China
Abstract:Beacon recognition is an important and critical technique in short range space rendezvous and docking. In the process of space rendezvous and docking, the imaging of CCD is affected by the interference of the sun, the moon and other kinds of celestial bodies as well as the radiation and reflection of the components of spacecraft, which forms a lot of miscellaneous lights. In order to eliminate the interference of all miscellaneous light and to complete the identification of the image points of each beacon, a kind of beacon recognition technique based on invariant features was designed. Firstly, a set of beacon system with invariant features was devised. Then, according to the inherent invariant features of beacon system, a set of complete algorithm was designed including pixel matching between left image and right image, the invariant feature recognition in far field and the forecasting recognition at near field in image plane as well as the model matching congruent recognition in image space. A lot of experiment results with imitation images and real images verify the correctness and reliability of our algorithms.
Keywords:Beacon recognition  Invariant features  Pixel matching  Forecasting recognition
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号