Multi-perspective Reconstruction

Abstract

       We develop new techniques for estimating scene geometry from a single or multiple images captured by multi-perspective cameras. Our priorities are given to efficient ray-geometry representations, novel stereo matching algorithms, and shape-from-distortion techniques.

Related Papers


Multi-perspective Stereo Matching and Volumetric Reconstruction
Yuanyuan Ding, Jingyi Yu, Peter Sturm
the 2009 International Conference on Computer Vision (ICCV '09)
 
 

Recovering Specular Surfaces Using Curved Line Images
Yuanyuan Ding, Jingyi Yu, Peter Sturm
in Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2009)
 
  State of the Art Report: Multiperspective Rendering, Modeling, and Imaging
Jingyi Yu, Leonard McMillan, and Peter Sturm
To Appear Eurographics 2008
 


Epsilon Stereo Pairs
Yuanyuan Ding and Jingyi Yu
To Appear British Machine Vision Conference (BMVC), 2007