Share on Facebook Tweet on Twitter Share on LinkedIn Share by email
Stereo Matching with Non-Linear Diffusion

Daniel Scharstein and Richard Szeliski

Abstract

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.

Details

Publication typeInproceedings
Published inIEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'96)
Pages343-350
AddressSan Francisco
PublisherIEEE Computer Society
> Publications > Stereo Matching with Non-Linear Diffusion