| 一种基于图象物理特征的分层匹配算法 |
| |
| 引用本文： | 杨小冈,缪栋,付光远. 一种基于图象物理特征的分层匹配算法[J]. 计算机工程与应用, 2002, 38(10): 106-108 |
| |
| 作者姓名： | 杨小冈 缪栋 付光远 |
| |
| 作者单位： | 第二炮兵工程学院,西安,710025 |
| |
| 摘 要： | 文章将物理学相关概念引入图象匹配中,定义了二维灰度图象的质量、密度、重心以及物理特征矢量,并由此提出了一种新的基于图象物理特征的分层匹配算法。该算法首先利用图象物理特征矢量进行粗匹配,找出几个大致的匹配点,再利用灰度归一化积相关算法进行精匹配。实验结果表明该算法具有较强的抗噪声干扰及几何畸变能力,与传统的归一化积相关算法相比,该算法在保证较高匹配概率的基础上,大大缩短了匹配时间,提高了匹配效率。
|
| 关 键 词： | 图象匹配 物理特征 归一化积相关 |
| 文章编号： | 1002-8331-(2002)10-0106-03 |
| 修稿时间： | 2001-04-01 |
A Hierarchical Matching Algorithm Based on Image Physical Feature |
| |
| Yang Xiaogang Miao Dong Fu Guangyuan. A Hierarchical Matching Algorithm Based on Image Physical Feature[J]. Computer Engineering and Applications, 2002, 38(10): 106-108 |
| |
| Authors: | Yang Xiaogang Miao Dong Fu Guangyuan |
| |
| Abstract: | Physics conceptions are introduced into image matching and definitions to the Mess,Density,Gravity Center and the Physical Feature Vector(PFV)of image are given in the paper.And a new matching algorithm based on image physical feature is proposed.It includes two steps.Firstly,the PFV of image is used for coarse matching.Then the Nor-malized Product correlation(NProd)algorithm is used for precise matching based on the first step's results.The experi-mental result demonstrates that the algorithm has the ability of resisting noise disturbs and geometric distortions and is more efficient compared with the NProd algorithm. |
| |
| Keywords: | Image matching Physical feature Normalized product correlation |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
