Fast line-segment extraction for semi-dense stereo matching

Brian McKinnon, Hansjoerg (Jacky) Baltes

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

Abstract

This paper describes our work on practical stereo vision for mobile robots using commodity hardware. The approach described in this paper is based on line segments, since those provide a lot of information about the environment, provide more depth information than point features, and are robust to image noise and colour variations. However, stereo matching with line segments is a difficult problem due to poorly localized end points and perspective distortion. Our algorithm uses integral images and Haar features for line segment extraction. Dynamic programming is used in the line segment matching phase. The resulting line segments track accurately from one frame to the next, even in the presence of noise.

Original languageEnglish
Title of host publicationRobot Vision - Second International Workshop, RobVis 2008, Proceedings
Pages59-71
Number of pages13
DOIs
Publication statusPublished - 2008 Aug 27
Event2nd International Workshop on Robot Vision, RobVis 2008 - University of Auckland, New Zealand
Duration: 2008 Feb 182008 Feb 20

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4931 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other2nd International Workshop on Robot Vision, RobVis 2008
CountryNew Zealand
CityUniversity of Auckland
Period08/2/1808/2/20

Fingerprint

Stereo Matching
Phase matching
Stereo vision
Line segment
Dynamic programming
Mobile robots
Color
Hardware
Stereo Vision
Feature Point
End point
Mobile Robot
Dynamic Programming

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

McKinnon, B., & Baltes, H. J. (2008). Fast line-segment extraction for semi-dense stereo matching. In Robot Vision - Second International Workshop, RobVis 2008, Proceedings (pp. 59-71). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4931 LNCS). https://doi.org/10.1007/978-3-540-78157-8_5

Fast line-segment extraction for semi-dense stereo matching. / McKinnon, Brian; Baltes, Hansjoerg (Jacky).

Robot Vision - Second International Workshop, RobVis 2008, Proceedings. 2008. p. 59-71 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4931 LNCS).

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

McKinnon, B & Baltes, HJ 2008, Fast line-segment extraction for semi-dense stereo matching. in Robot Vision - Second International Workshop, RobVis 2008, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4931 LNCS, pp. 59-71, 2nd International Workshop on Robot Vision, RobVis 2008, University of Auckland, New Zealand, 08/2/18. https://doi.org/10.1007/978-3-540-78157-8_5
McKinnon B, Baltes HJ. Fast line-segment extraction for semi-dense stereo matching. In Robot Vision - Second International Workshop, RobVis 2008, Proceedings. 2008. p. 59-71. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-540-78157-8_5
McKinnon, Brian ; Baltes, Hansjoerg (Jacky). / Fast line-segment extraction for semi-dense stereo matching. Robot Vision - Second International Workshop, RobVis 2008, Proceedings. 2008. pp. 59-71 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{92248d746f89417eb07fa9299f1f5c92,
title = "Fast line-segment extraction for semi-dense stereo matching",
abstract = "This paper describes our work on practical stereo vision for mobile robots using commodity hardware. The approach described in this paper is based on line segments, since those provide a lot of information about the environment, provide more depth information than point features, and are robust to image noise and colour variations. However, stereo matching with line segments is a difficult problem due to poorly localized end points and perspective distortion. Our algorithm uses integral images and Haar features for line segment extraction. Dynamic programming is used in the line segment matching phase. The resulting line segments track accurately from one frame to the next, even in the presence of noise.",
author = "Brian McKinnon and Baltes, {Hansjoerg (Jacky)}",
year = "2008",
month = "8",
day = "27",
doi = "10.1007/978-3-540-78157-8_5",
language = "English",
isbn = "3540781560",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "59--71",
booktitle = "Robot Vision - Second International Workshop, RobVis 2008, Proceedings",

}

TY - GEN

T1 - Fast line-segment extraction for semi-dense stereo matching

AU - McKinnon, Brian

AU - Baltes, Hansjoerg (Jacky)

PY - 2008/8/27

Y1 - 2008/8/27

N2 - This paper describes our work on practical stereo vision for mobile robots using commodity hardware. The approach described in this paper is based on line segments, since those provide a lot of information about the environment, provide more depth information than point features, and are robust to image noise and colour variations. However, stereo matching with line segments is a difficult problem due to poorly localized end points and perspective distortion. Our algorithm uses integral images and Haar features for line segment extraction. Dynamic programming is used in the line segment matching phase. The resulting line segments track accurately from one frame to the next, even in the presence of noise.

AB - This paper describes our work on practical stereo vision for mobile robots using commodity hardware. The approach described in this paper is based on line segments, since those provide a lot of information about the environment, provide more depth information than point features, and are robust to image noise and colour variations. However, stereo matching with line segments is a difficult problem due to poorly localized end points and perspective distortion. Our algorithm uses integral images and Haar features for line segment extraction. Dynamic programming is used in the line segment matching phase. The resulting line segments track accurately from one frame to the next, even in the presence of noise.

UR - http://www.scopus.com/inward/record.url?scp=49949101926&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=49949101926&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-78157-8_5

DO - 10.1007/978-3-540-78157-8_5

M3 - Conference contribution

AN - SCOPUS:49949101926

SN - 3540781560

SN - 9783540781561

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

SP - 59

EP - 71

BT - Robot Vision - Second International Workshop, RobVis 2008, Proceedings

ER -