Stereo Matching with Nonlinear Diffusion

被引：0
作者：
Daniel Scharstein
Richard Szeliski
机构：
[1] Middlebury College,Department of Mathematics and Computer Science
[2] One Microsoft Way,Microsoft Research
来源：
International Journal of Computer Vision | 1998年 / 28卷
关键词：
stereo matching; variable-sized support region; nonlinear diffusion; Bayesian estimation;
D O I：
暂无
中图分类号：
学科分类号：
摘要：
One of the central problems in stereo matching (and other image registration tasks) is the selection of optimal window sizes for comparing image regions. This paper addresses this problem with some novel algorithms based on iteratively diffusing support at different disparity hypotheses, and locally controlling the amount of diffusion based on the current quality of the disparity estimate. It also develops a novel Bayesian estimation technique, which significantly outperforms techniques based on area-based matching (SSD) and regular diffusion. We provide experimental results on both synthetic and real stereo image pairs.
引用
收藏
页码：155 / 174
页数：19
相关论文
共 74 条
[21]  
Bobick A.F.(1993)A multiple-baseline stereo IEEE Transactions on Pattern Analysis and Machine Intelligence 15 353-363
[22]  
Jenkin M.R.M.(1990)Stereo correspondence by surface reconstruction IEEE Transactions on Pattern Analysis and Machine Intelligence 12 309-314
[23]  
Jepson A.D.(1990)Scale-space and edge detection using anisotropic diffusion IEEE Transactions on Pattern Analysis and Machine Intelligence 12 629-639
[24]  
Tsotsos J.K.(1989)A new method of mapping optimization problems onto neural networks International Journal of Neural Systems 1 3-470
[25]  
Kanade T.(1985)PMF: A stereo correspondence algorithm using a disparity gradient limit Perception 14 449-99
[26]  
Okutomi M.(1985)Detection of binocular disparities Biological Cybernetics 52 93-304
[27]  
Marr D.(1994)Determination of optical flow and its discontinuities using nonlinear diffusion Third European Conference on Computer Vision (ECCV’94), Stockholm, Sweden 2 295-322
[28]  
Poggio T.(1980)Prediction of correlation errors in stereo-pair images Optical Engineering 19 312-575
[29]  
Marr D.C.(1994)Matching images by comparing their gradient fields 12th International Conference on Pattern Recognition (ICPR’94), Jerusalem, Israel 1 572-1639
[30]  
Poggio T.(1989)Using local orientation information as image primitive for robust object recognition SPIE Visual Communications and Image Processing IV 1199 1630-168
12345678