Title :
Quasi-dense 3D reconstruction using tensor-based multiview stereo
Author :
Wu, Tai-Pang ; Yeung, Sai-Kit ; Jia, Jiaya ; Tang, Chi-Keung
Author_Institution :
Hong Kong Univ. of Sci. & Technol., Hong Kong, China
Abstract :
We propose tensor-based multiview stereo (TMVS) for quasi-dense 3D reconstruction from uncalibrated images. Our work is inspired by the patch-based multiview stereo (PMVS), a state-of-the-art technique in multiview stereo reconstruction. The effectiveness of PMVS is attributed to the use of 3D patches in the match-propagate-filter MVS pipeline. Our key observation is: PMVS has not fully utilized the valuable 3D geometric cue available in 3D patches which are oriented points. This paper combines the complementary advantages of photoconsistency, visibility and geometric consistency enforcement in MVS via the use of 3D tensors, where our closed-form solution to tensor voting provides a unified approach to implement the match-propagate-filter pipeline. Using PMVS as the implementation backbone where TMVS is built, we provide qualitative and quantitative evaluation to demonstrate how TMVS significantly improve the MVS pipeline.
Keywords :
computational geometry; filtering theory; image reconstruction; matched filters; stereo image processing; tensors; video signal processing; 3D geometric cue; 3D patches; geometric consistency enforcement; match-propagate-filter MVS pipeline; multiview stereo reconstruction; patch-based multiview stereo; photoconsistency; quasi-dense 3D reconstruction; tensor voting; tensor-based multiview stereo; uncalibrated images; Closed-form solution; Image reconstruction; Information geometry; Pipelines; Smoothing methods; Spine; Stereo image processing; Surface reconstruction; Tensile stress; Voting;
Conference_Titel :
Computer Vision and Pattern Recognition (CVPR), 2010 IEEE Conference on
Conference_Location :
San Francisco, CA
Print_ISBN :
978-1-4244-6984-0
DOI :
10.1109/CVPR.2010.5539796