Robust stereo matching using probabilistic laplacian surface propagation

Seungryong Kim; Bumsub Ham; Seungchul Ryu; Seon Joo Kim; Kwanghoon Sohn

doi:10.1007/978-3-319-16865-4_24

Robust stereo matching using probabilistic laplacian surface propagation

Seungryong Kim, Bumsub Ham, Seungchul Ryu, Seon Joo Kim, Kwanghoon Sohn

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

3 Citations (Scopus)

Abstract

This paper describes a probabilistic Laplacian surface propagation (PLSP) framework for a robust stereo matching under severe radiometric variations. We discover that a progressive scheme overcomes an inherent limitation for this task, while most conventional efforts have been focusing on designing a robust cost function. We propose the ground control surfaces (GCSs) designed as progressive unit, which alleviates the problems of conventional progressive methods and superpixel based methods, simultaneously. Moreover, we introduce a novel confidence measure for stereo pairs taken under radiometric variations based on the probability of correspondences. Specifically, the PLSP estimates the GCSs from initial sparse disparity maps using a weighted least-square. The GCSs are then propagated on a superpixel graph with a surface confidence weighting. Experimental results show that the PLSP outperforms state-of-the-art robust cost function based methods and other propagation methods for the stereo matching under radiometric variations.

Original language	English
Title of host publication	Computer Vision - ACCV 2014 - 12th Asian Conference on Computer Vision, Revised Selected Papers
Editors	Ian Reid, Ming-Hsuan Yang, Hideo Saito, Daniel Cremers
Publisher	Springer Verlag
Pages	368-383
Number of pages	16
ISBN (Electronic)	9783319168647
DOIs	https://doi.org/10.1007/978-3-319-16865-4_24
Publication status	Published - 2015
Event	12th Asian Conference on Computer Vision, ACCV 2014 - Singapore, Singapore Duration: 2014 Nov 1 → 2014 Nov 5

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9003
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	12th Asian Conference on Computer Vision, ACCV 2014
Country/Territory	Singapore
City	Singapore
Period	14/11/1 → 14/11/5

Bibliographical note

Funding Information:
This research was supported by the MSIP(Ministry of Science, ICT and Future Planning), Korea, under the ITRC(Information Technology Research Center) support program (NIPA-2014-H0301-14-1012) supervised by the NIPA(National IT Industry Promotion Agency).

Publisher Copyright:
© Springer International Publishing Switzerland 2015.

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-319-16865-4_24

Cite this

Kim, S., Ham, B., Ryu, S., Kim, S. J., & Sohn, K. (2015). Robust stereo matching using probabilistic laplacian surface propagation. In I. Reid, M-H. Yang, H. Saito, & D. Cremers (Eds.), Computer Vision - ACCV 2014 - 12th Asian Conference on Computer Vision, Revised Selected Papers (pp. 368-383). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9003). Springer Verlag. https://doi.org/10.1007/978-3-319-16865-4_24

Kim, Seungryong ; Ham, Bumsub ; Ryu, Seungchul et al. / Robust stereo matching using probabilistic laplacian surface propagation. Computer Vision - ACCV 2014 - 12th Asian Conference on Computer Vision, Revised Selected Papers. editor / Ian Reid ; Ming-Hsuan Yang ; Hideo Saito ; Daniel Cremers. Springer Verlag, 2015. pp. 368-383 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{e07bb580d339433aaf75d63f5b29448d,

title = "Robust stereo matching using probabilistic laplacian surface propagation",

abstract = "This paper describes a probabilistic Laplacian surface propagation (PLSP) framework for a robust stereo matching under severe radiometric variations. We discover that a progressive scheme overcomes an inherent limitation for this task, while most conventional efforts have been focusing on designing a robust cost function. We propose the ground control surfaces (GCSs) designed as progressive unit, which alleviates the problems of conventional progressive methods and superpixel based methods, simultaneously. Moreover, we introduce a novel confidence measure for stereo pairs taken under radiometric variations based on the probability of correspondences. Specifically, the PLSP estimates the GCSs from initial sparse disparity maps using a weighted least-square. The GCSs are then propagated on a superpixel graph with a surface confidence weighting. Experimental results show that the PLSP outperforms state-of-the-art robust cost function based methods and other propagation methods for the stereo matching under radiometric variations.",

author = "Seungryong Kim and Bumsub Ham and Seungchul Ryu and Kim, {Seon Joo} and Kwanghoon Sohn",

note = "Funding Information: This research was supported by the MSIP(Ministry of Science, ICT and Future Planning), Korea, under the ITRC(Information Technology Research Center) support program (NIPA-2014-H0301-14-1012) supervised by the NIPA(National IT Industry Promotion Agency). Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2015.; 12th Asian Conference on Computer Vision, ACCV 2014 ; Conference date: 01-11-2014 Through 05-11-2014",

year = "2015",

doi = "10.1007/978-3-319-16865-4_24",

language = "English",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "368--383",

editor = "Ian Reid and Ming-Hsuan Yang and Hideo Saito and Daniel Cremers",

booktitle = "Computer Vision - ACCV 2014 - 12th Asian Conference on Computer Vision, Revised Selected Papers",

address = "Germany",

}

Kim, S, Ham, B, Ryu, S, Kim, SJ & Sohn, K 2015, Robust stereo matching using probabilistic laplacian surface propagation. in I Reid, M-H Yang, H Saito & D Cremers (eds), Computer Vision - ACCV 2014 - 12th Asian Conference on Computer Vision, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9003, Springer Verlag, pp. 368-383, 12th Asian Conference on Computer Vision, ACCV 2014, Singapore, Singapore, 14/11/1. https://doi.org/10.1007/978-3-319-16865-4_24

Robust stereo matching using probabilistic laplacian surface propagation. / Kim, Seungryong; Ham, Bumsub; Ryu, Seungchul et al.
Computer Vision - ACCV 2014 - 12th Asian Conference on Computer Vision, Revised Selected Papers. ed. / Ian Reid; Ming-Hsuan Yang; Hideo Saito; Daniel Cremers. Springer Verlag, 2015. p. 368-383 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9003).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Robust stereo matching using probabilistic laplacian surface propagation

AU - Kim, Seungryong

AU - Ham, Bumsub

AU - Ryu, Seungchul

AU - Kim, Seon Joo

AU - Sohn, Kwanghoon

N1 - Funding Information: This research was supported by the MSIP(Ministry of Science, ICT and Future Planning), Korea, under the ITRC(Information Technology Research Center) support program (NIPA-2014-H0301-14-1012) supervised by the NIPA(National IT Industry Promotion Agency). Publisher Copyright: © Springer International Publishing Switzerland 2015.

PY - 2015

Y1 - 2015

N2 - This paper describes a probabilistic Laplacian surface propagation (PLSP) framework for a robust stereo matching under severe radiometric variations. We discover that a progressive scheme overcomes an inherent limitation for this task, while most conventional efforts have been focusing on designing a robust cost function. We propose the ground control surfaces (GCSs) designed as progressive unit, which alleviates the problems of conventional progressive methods and superpixel based methods, simultaneously. Moreover, we introduce a novel confidence measure for stereo pairs taken under radiometric variations based on the probability of correspondences. Specifically, the PLSP estimates the GCSs from initial sparse disparity maps using a weighted least-square. The GCSs are then propagated on a superpixel graph with a surface confidence weighting. Experimental results show that the PLSP outperforms state-of-the-art robust cost function based methods and other propagation methods for the stereo matching under radiometric variations.

AB - This paper describes a probabilistic Laplacian surface propagation (PLSP) framework for a robust stereo matching under severe radiometric variations. We discover that a progressive scheme overcomes an inherent limitation for this task, while most conventional efforts have been focusing on designing a robust cost function. We propose the ground control surfaces (GCSs) designed as progressive unit, which alleviates the problems of conventional progressive methods and superpixel based methods, simultaneously. Moreover, we introduce a novel confidence measure for stereo pairs taken under radiometric variations based on the probability of correspondences. Specifically, the PLSP estimates the GCSs from initial sparse disparity maps using a weighted least-square. The GCSs are then propagated on a superpixel graph with a surface confidence weighting. Experimental results show that the PLSP outperforms state-of-the-art robust cost function based methods and other propagation methods for the stereo matching under radiometric variations.

UR - http://www.scopus.com/inward/record.url?scp=84938860671&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84938860671&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-16865-4_24

DO - 10.1007/978-3-319-16865-4_24

M3 - Conference contribution

AN - SCOPUS:84938860671

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 368

EP - 383

BT - Computer Vision - ACCV 2014 - 12th Asian Conference on Computer Vision, Revised Selected Papers

A2 - Reid, Ian

A2 - Yang, Ming-Hsuan

A2 - Saito, Hideo

A2 - Cremers, Daniel

PB - Springer Verlag

T2 - 12th Asian Conference on Computer Vision, ACCV 2014

Y2 - 1 November 2014 through 5 November 2014

ER -

Kim S, Ham B, Ryu S, Kim SJ , Sohn K. Robust stereo matching using probabilistic laplacian surface propagation. In Reid I, Yang M-H, Saito H, Cremers D, editors, Computer Vision - ACCV 2014 - 12th Asian Conference on Computer Vision, Revised Selected Papers. Springer Verlag. 2015. p. 368-383. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-16865-4_24

Robust stereo matching using probabilistic laplacian surface propagation

Abstract

Publication series

Other

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this