Robust stereo matching using adaptive random walk with restart algorithm

作者：

Highlights：

• We develop a robust stereo matching algorithm using a random walk with restart.

• The proposed method considers occlusion and depth discontinuities.

• The method achieves high accuracy with low computational costs in the reconstruction.

• Our method works well on the varying illumination and exposure test.

摘要

•We develop a robust stereo matching algorithm using a random walk with restart.•The proposed method considers occlusion and depth discontinuities.•The method achieves high accuracy with low computational costs in the reconstruction.•Our method works well on the varying illumination and exposure test.

论文关键词：Global optimization,Random walk with restart,Stereo matching,Superpixels

论文评审过程：Received 9 November 2013, Revised 4 October 2014, Accepted 9 January 2015, Available online 20 February 2015.

论文官网地址：https://doi.org/10.1016/j.imavis.2015.01.003