全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

基于网络最小割的分层立体视觉匹配方法*

, PP. 64-68

Keywords: 立体匹配,视差,最小割,区域相关

Full-Text   Cite this paper   Add to My Lib

Abstract:

立体匹配是计算机视觉领域中的一个难点问题.为了得到准确的高密度的视差图,本文提出一种基于网络最小割的分层匹配方法.该方法综合运用区域灰度相关法和最小割全局最优搜索策略.首先对原图像对进行两层金字塔分解,在低分辨率的图像中运用网络最小割方法求得全局最优匹配.然后在低分辨率的图像中匹配的像素对的约束下,在原图像对中采用区域灰度相关法进行匹配,得到高密度视差图.这样既缩小匹配时的搜索空间,又保证匹配的可靠性.实验表明,该方法是有效可行的.

References

[1]  Scharstein D, Szeliski R. A Taxonomy and Evaluation of Dense TwoFrame Stereo Correspondence Algorithms. International Journal of Computer Vision, 2002, 47(1/2/3): 742
[2]  Crespi B, Cozzi A G. Analog Computation for PhaseBased Disparity Estimation: Continuous and Discrete Models. Machine Vision and Applications, 1998, 11(2): 8395
[3]  Greig D M, Porteous B D, Seheult A H. Exact Maximum a Posteriori Estimation for Binary Image. Journal of the Royal Statistical Society, 1989, 51(2): 271279
[4]  Roy S. Stereo without Epipolar Lines: A MaximumFlow Formulation. International Journal of Computer Vision, 1999, 34(2/3): 147161
[5]  Boykov Y, Veksler O, Zabih R. Fast Approximate Energy Minimization via Graph Cuts. IEEE Trans on Pattern Analysis and Machine Intelligence, 2001, 23(11): 12221239
[6]  Boykov Y, Kolmogorov V. An Experimental Comparison of MinCut/MaxFlow Algorithms for Energy Minimization in Vision. IEEE Trans on Pattern Analysis and Machine Intelligence, 2004, 26(9): 11241137
[7]  Huang Haiyun, Qi Feihu, Cen Feng. Stereo Matching Algorithm Based on the Maximum Flow of Network. Journal of Shanghai Jiaotong University, 2001, 35(2): 168172 (in Chinese) (黄海赟,戚飞虎,岑 峰.基于网络最大流的立体匹配算法. 上海交通大学学报, 2001, 35(2): 168172)
[8]  Xie Zheng. Network Algorithm and Complexity Theory. 2nd Edition. Changsha, China: National Defense Science and Technology University Press, 2003 (in Chinese) (谢 政.网络算法与复杂性理论.第2版.长沙:国防科技大学出版社, 2003)
[9]  Cochran S D, Medioni G. 3D Surface Description from Binocular Stereo. IEEE Trans on Pattern Analysis and Machine Intelligence, 1992, 14(10): 981994

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133