Journal Home Online First Current Issue Archive For Authors Journal Information 中文版

Strategic Study of CAE >> 2010, Volume 12, Issue 1

A algorithm for binocular stereo matching in the presence of specular reflections

Department of Computer Science & Technology, Nanjing University of Science and Technology, Nanjing 210094, China

Funding project:国家部委基金资助项目(51316080101) Received: 2007-12-24 Available online: 2010-01-14 13:32:38.000

Next Previous

Abstract

Traditional stereo correspondence algorithms rely heavily on the Lambertian model of diffuse reflectance. While this diffuse assumption is generally valid for most part of an image, processing of regions that contain specular reflections can result in severe matching errors. In this paper, we address the problem of binocular stereo dense matching in the presence of specular reflections by introducing a novel correspondence measurement which is robust to the specular reflections. Accurate depth can be estimated for both diffuse and specular regions. Unlike the previous works which seek to eliminate or avoid specular reflections using image preprocessing or multibaseline stereo, our approach works in its presence. Experiments demonstrate the effectiveness and robustness of our approach.

Figures

图1

图2

图3

图4

图5

References

[ 1 ] Zhou Wei,Kambhamettu C .Binocular Stereo Dense Matching in the Presence of Specular Reflections [ A ] .proceedings of 2006 IEEE computer society conference on computer vision and pattern Recognition [ C] .New Yoke.2006 ,2 :2363 -2370

[ 2 ] Torrance K , Sparrow E.Theory for off -specular reflection from roughened surfaces[ J] .Journal of the Optics Society of America, 1967 ,57 ( 9 ) :1105 -1114 link1

[ 3 ] Yang Yibing, Alan Yuille, Lu Jie.Local, global and multilevel stereo matching[ A] .Proceedings of 2006 IEEE computer society conference on computer vision and pattern Recognition [ C] .New York.1993 :274 -279

[ 4 ] Shafer S.Using color to separate reflection Components[ J] .Color Research and Applications, 1985 ,10 ( 4 ) :210 -218 link1

Related Research