Reconstruction of 3D models consisting of line segments

Naoto Ienaga, Hideo Saito

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

1 Citation (Scopus)

Abstract

Reconstruction of three-dimensional (3D) models from images is currently one of the most important areas of research in computer vision. In this paper, we propose a method to recover 3D models using the minimum number of line segments. By using structure-from-motion, the proposed method first recovers a 3D model of line segments detected from an input image sequence. We then detect overlapping 3D line segments that redundantly represent a single line structure so that the number of 3D line segments representing the target scene can be reduced without losing the detailed geometry of the structure. We apply matching and depth information to remove redundant line segments from the model while keeping the necessary segments. In experiments, we confirm that the proposed method can greatly reduce the number of line segments. We also demonstrate that the accuracy and computational time for camera pose estimation can be significantly improved with the 3D line segment model recovered by the proposed method. Moreover, we have applied the proposed method to see through occluded areas.

Original languageEnglish
Title of host publicationComputer Vision - ACCV 2016 Workshops, ACCV 2016 International Workshops, Revised Selected Papers
PublisherSpringer Verlag
Pages100-113
Number of pages14
Volume10117 LNCS
ISBN (Print)9783319544267
DOIs
Publication statusPublished - 2017
Event13th Asian Conference on Computer Vision, ACCV 2016 - Taipei, Taiwan, Province of China
Duration: 2016 Nov 202016 Nov 24

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10117 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other13th Asian Conference on Computer Vision, ACCV 2016
CountryTaiwan, Province of China
City Taipei
Period16/11/2016/11/24

Fingerprint

Line segment
3D Model
Structure from Motion
Computer vision
Pose Estimation
Cameras
Image Sequence
Computer Vision
Overlapping
Geometry
Camera
Three-dimensional
Target
Necessary
Line
Experiments
Model
Demonstrate
Experiment

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Ienaga, N., & Saito, H. (2017). Reconstruction of 3D models consisting of line segments. In Computer Vision - ACCV 2016 Workshops, ACCV 2016 International Workshops, Revised Selected Papers (Vol. 10117 LNCS, pp. 100-113). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10117 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-54427-4_8

Reconstruction of 3D models consisting of line segments. / Ienaga, Naoto; Saito, Hideo.

Computer Vision - ACCV 2016 Workshops, ACCV 2016 International Workshops, Revised Selected Papers. Vol. 10117 LNCS Springer Verlag, 2017. p. 100-113 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10117 LNCS).

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

Ienaga, N & Saito, H 2017, Reconstruction of 3D models consisting of line segments. in Computer Vision - ACCV 2016 Workshops, ACCV 2016 International Workshops, Revised Selected Papers. vol. 10117 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10117 LNCS, Springer Verlag, pp. 100-113, 13th Asian Conference on Computer Vision, ACCV 2016, Taipei, Taiwan, Province of China, 16/11/20. https://doi.org/10.1007/978-3-319-54427-4_8
Ienaga N, Saito H. Reconstruction of 3D models consisting of line segments. In Computer Vision - ACCV 2016 Workshops, ACCV 2016 International Workshops, Revised Selected Papers. Vol. 10117 LNCS. Springer Verlag. 2017. p. 100-113. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-54427-4_8
Ienaga, Naoto ; Saito, Hideo. / Reconstruction of 3D models consisting of line segments. Computer Vision - ACCV 2016 Workshops, ACCV 2016 International Workshops, Revised Selected Papers. Vol. 10117 LNCS Springer Verlag, 2017. pp. 100-113 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{912fdcc9598a454ca42a414e1c5c2f31,
title = "Reconstruction of 3D models consisting of line segments",
abstract = "Reconstruction of three-dimensional (3D) models from images is currently one of the most important areas of research in computer vision. In this paper, we propose a method to recover 3D models using the minimum number of line segments. By using structure-from-motion, the proposed method first recovers a 3D model of line segments detected from an input image sequence. We then detect overlapping 3D line segments that redundantly represent a single line structure so that the number of 3D line segments representing the target scene can be reduced without losing the detailed geometry of the structure. We apply matching and depth information to remove redundant line segments from the model while keeping the necessary segments. In experiments, we confirm that the proposed method can greatly reduce the number of line segments. We also demonstrate that the accuracy and computational time for camera pose estimation can be significantly improved with the 3D line segment model recovered by the proposed method. Moreover, we have applied the proposed method to see through occluded areas.",
author = "Naoto Ienaga and Hideo Saito",
year = "2017",
doi = "10.1007/978-3-319-54427-4_8",
language = "English",
isbn = "9783319544267",
volume = "10117 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "100--113",
booktitle = "Computer Vision - ACCV 2016 Workshops, ACCV 2016 International Workshops, Revised Selected Papers",
address = "Germany",

}

TY - GEN

T1 - Reconstruction of 3D models consisting of line segments

AU - Ienaga, Naoto

AU - Saito, Hideo

PY - 2017

Y1 - 2017

N2 - Reconstruction of three-dimensional (3D) models from images is currently one of the most important areas of research in computer vision. In this paper, we propose a method to recover 3D models using the minimum number of line segments. By using structure-from-motion, the proposed method first recovers a 3D model of line segments detected from an input image sequence. We then detect overlapping 3D line segments that redundantly represent a single line structure so that the number of 3D line segments representing the target scene can be reduced without losing the detailed geometry of the structure. We apply matching and depth information to remove redundant line segments from the model while keeping the necessary segments. In experiments, we confirm that the proposed method can greatly reduce the number of line segments. We also demonstrate that the accuracy and computational time for camera pose estimation can be significantly improved with the 3D line segment model recovered by the proposed method. Moreover, we have applied the proposed method to see through occluded areas.

AB - Reconstruction of three-dimensional (3D) models from images is currently one of the most important areas of research in computer vision. In this paper, we propose a method to recover 3D models using the minimum number of line segments. By using structure-from-motion, the proposed method first recovers a 3D model of line segments detected from an input image sequence. We then detect overlapping 3D line segments that redundantly represent a single line structure so that the number of 3D line segments representing the target scene can be reduced without losing the detailed geometry of the structure. We apply matching and depth information to remove redundant line segments from the model while keeping the necessary segments. In experiments, we confirm that the proposed method can greatly reduce the number of line segments. We also demonstrate that the accuracy and computational time for camera pose estimation can be significantly improved with the 3D line segment model recovered by the proposed method. Moreover, we have applied the proposed method to see through occluded areas.

UR - http://www.scopus.com/inward/record.url?scp=85016095067&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85016095067&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-54427-4_8

DO - 10.1007/978-3-319-54427-4_8

M3 - Conference contribution

AN - SCOPUS:85016095067

SN - 9783319544267

VL - 10117 LNCS

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

SP - 100

EP - 113

BT - Computer Vision - ACCV 2016 Workshops, ACCV 2016 International Workshops, Revised Selected Papers

PB - Springer Verlag

ER -