首页 | 本学科首页   官方微博 | 高级检索  
     

一种高精度的图像匹配算法
引用本文:王兆仲,周付根,刘志芳,杨建峰.一种高精度的图像匹配算法[J].红外与激光工程,2006,35(6):751-755.
作者姓名:王兆仲  周付根  刘志芳  杨建峰
作者单位:1. 北京航空航天大学,图像中心,北京,100083
2. 中国科学院西安光学与精密机械研究所,陕西,西安,710068
基金项目:国家高技术研究发展计划(863计划)
摘    要:为了解决光谱恢复对图像匹配高精度的问题,提出了一种高精度图像匹配算法——利用光流确定图像的运动场对图像进行匹配。该算法克服了传统的基于灰度匹配方法受图像插值精度影响的缺陷以及运算速度和精度的局限性。与基于MAD块匹配算法和归一化相关系数块匹配算法进行比较,该算法实现了图像序列高精度匹配,而且硬件实现简单,计算复杂度较低。

关 键 词:图像匹配    光流    高光谱图像    运动估计
文章编号:1007-2276(2006)06-0751-05
收稿时间:2005-12-10
修稿时间:2006-03-08

Image matching algorithm with high precision
WANG Zhao-zhong,ZHOU Fu-gen,LIU Zhi-fang,YANG Jian-feng.Image matching algorithm with high precision[J].Infrared and Laser Engineering,2006,35(6):751-755.
Authors:WANG Zhao-zhong  ZHOU Fu-gen  LIU Zhi-fang  YANG Jian-feng
Affiliation:1.Image Processing Center, Beihang University,Beijing 100083, China; 2.Xi′an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences,Xi′an 710068, China
Abstract:A novel algorithm for image matching with high accuracy is proposed to achieve the requirement of spectrum restoration based on optical flow computation. This matching algorithm overcomes the defects of traditional image matching algorithm based on gray matching, which are affected by interpolation accuracy, and breaks the conflict between operation speed and accuracy. Experiments on real hyper-spectral image sequences show that our method is not only effective both in efficiency and accuracy, but also more suitable for the hardware realization compared with MAD block matching algorithm and normalized correlation coefficient block matching algorithm.
Keywords:Image matching  Optical flow  Hyper-spectral image  Motion estimation
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《红外与激光工程》浏览原始摘要信息
点击此处可从《红外与激光工程》下载全文
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