基于边缘特征点的分级立体匹配算法研究-陕西师范大学学报期刊社网站

陕西师范大学学报（自然科学版）

数学与计算机科学

基于边缘特征点的分级立体匹配算法研究

李刚1，李竹林2，赵宗涛1

((1 西北大学信息科学与技术学院，陕西西安 710069； 2 延安大学数学与计算机科学学院，陕西延安 716000)

李刚，男，博士研究生，研究方向为计算机软件与理论．

摘要:

提出了一种基于边缘特征点的分级立体匹配算法．其基本思想是，首先提取双目图像对的边缘特征点，然后对特征点的梯度不变性与奇异值特征不变性进行分析，建立基于不变量特征的二级匹配算法，求解基础矩阵．在基础矩阵的指导下，完成三级立体匹配．模拟结果表明，该算法能使匹配精度由58.3%提高到73.2%，而且算法简单、实用，对目标的识别、跟踪以及三维表面恢复与重建等均具有重要的价值．

关键词：

边缘特征点；立体匹配；梯度；奇异特征值

收稿日期：

2009-03-26

中图分类号：

TP391.41

文献标识码：

文章编号：

1672-4291(2009)05-0020-04

基金项目：

中国博士后基金资助项目(080431401)

Doi:

Research on gradation stereo matching algorithm based on the edge feature points

LI Gang1, LI Zhu-lin2, ZHAO Zong-tao1

(1 College of Information Sicence and Technology, Northwest University, Xi′an 710069, Shaanxi, China; 2 College of Mathematics and Computer Science, Yan′an University, Yan′an 716000, Shaanxi, China)

Abstract:

A gradation stereo matching algorithm based on the edge feature points is proposed. Its basic idea is that, the edge feature points of a pair of images are first extracted, and then gradient invariability and singular eigenvalue invariability are analyzed, two-grade stereo matching method is build, foundation matrix is solved further, and three-grade stereo matching algorithm is finished by foundation matrix guidance. The result indicates that the algorithm can improve matching precision, from 58.3% to 73.2%, it is simple and utility, and it is important for object recognition, tracking, and three-dimension reconstruction.

KeyWords:

edge feature point; stereo matching; gradient; singular eigenvalue