Shape from Light Field meets Robust PCA

Stefan Heber, Thomas Pock

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper we propose a new type of matching term for multi-view stereo reconstruction. Our model is based on the assumption, that if one warps the images of the various views to a common warping-center and considers each warped image as one row in a matrix, then this matrix will have low rank. This also implies, that we assume a certain amount of overlap between the views
after the warping has been performed. Such an assumption is obviously met in the case of light field data, which motivated us to demonstrate the proposed model for this type of data. Our final model is a large scale convex optimization problem, where the low rank minimization is relaxed via the nuclear norm. We present qualitative and quantitative experiments, where the proposed model achieves excellent results.
LanguageEnglish
Title of host publicationComputer Vision -- ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VI
PublisherSpringer International Publishing AG
Pages751-767
Volume8694
ISBN (Electronic)978-3-319-10599-4
ISBN (Print)978-3-319-10598-7
DOIs
StatusAccepted/In press - 2014

Fingerprint

Convex optimization
Experiments

Fields of Expertise

  • Information, Communication & Computing

Cite this

Heber, S., & Pock, T. (2014). Shape from Light Field meets Robust PCA. In Computer Vision -- ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VI (Vol. 8694, pp. 751-767). Springer International Publishing AG . DOI: 10.1007/978-3-319-10599-4_48

Shape from Light Field meets Robust PCA. / Heber, Stefan; Pock, Thomas.

Computer Vision -- ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VI. Vol. 8694 Springer International Publishing AG , 2014. p. 751-767.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Heber, S & Pock, T 2014, Shape from Light Field meets Robust PCA. in Computer Vision -- ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VI. vol. 8694, Springer International Publishing AG , pp. 751-767. DOI: 10.1007/978-3-319-10599-4_48
Heber S, Pock T. Shape from Light Field meets Robust PCA. In Computer Vision -- ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VI. Vol. 8694. Springer International Publishing AG . 2014. p. 751-767. Available from, DOI: 10.1007/978-3-319-10599-4_48
Heber, Stefan ; Pock, Thomas. / Shape from Light Field meets Robust PCA. Computer Vision -- ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VI. Vol. 8694 Springer International Publishing AG , 2014. pp. 751-767
@inproceedings{4de2a6a0cf4f4e55ac061556753d9334,
title = "Shape from Light Field meets Robust PCA",
abstract = "In this paper we propose a new type of matching term for multi-view stereo reconstruction. Our model is based on the assumption, that if one warps the images of the various views to a common warping-center and considers each warped image as one row in a matrix, then this matrix will have low rank. This also implies, that we assume a certain amount of overlap between the viewsafter the warping has been performed. Such an assumption is obviously met in the case of light field data, which motivated us to demonstrate the proposed model for this type of data. Our final model is a large scale convex optimization problem, where the low rank minimization is relaxed via the nuclear norm. We present qualitative and quantitative experiments, where the proposed model achieves excellent results.",
author = "Stefan Heber and Thomas Pock",
year = "2014",
doi = "10.1007/978-3-319-10599-4_48",
language = "English",
isbn = "978-3-319-10598-7",
volume = "8694",
pages = "751--767",
booktitle = "Computer Vision -- ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VI",
publisher = "Springer International Publishing AG",
address = "Switzerland",

}

TY - GEN

T1 - Shape from Light Field meets Robust PCA

AU - Heber,Stefan

AU - Pock,Thomas

PY - 2014

Y1 - 2014

N2 - In this paper we propose a new type of matching term for multi-view stereo reconstruction. Our model is based on the assumption, that if one warps the images of the various views to a common warping-center and considers each warped image as one row in a matrix, then this matrix will have low rank. This also implies, that we assume a certain amount of overlap between the viewsafter the warping has been performed. Such an assumption is obviously met in the case of light field data, which motivated us to demonstrate the proposed model for this type of data. Our final model is a large scale convex optimization problem, where the low rank minimization is relaxed via the nuclear norm. We present qualitative and quantitative experiments, where the proposed model achieves excellent results.

AB - In this paper we propose a new type of matching term for multi-view stereo reconstruction. Our model is based on the assumption, that if one warps the images of the various views to a common warping-center and considers each warped image as one row in a matrix, then this matrix will have low rank. This also implies, that we assume a certain amount of overlap between the viewsafter the warping has been performed. Such an assumption is obviously met in the case of light field data, which motivated us to demonstrate the proposed model for this type of data. Our final model is a large scale convex optimization problem, where the low rank minimization is relaxed via the nuclear norm. We present qualitative and quantitative experiments, where the proposed model achieves excellent results.

U2 - 10.1007/978-3-319-10599-4_48

DO - 10.1007/978-3-319-10599-4_48

M3 - Conference contribution

SN - 978-3-319-10598-7

VL - 8694

SP - 751

EP - 767

BT - Computer Vision -- ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VI

PB - Springer International Publishing AG

ER -