Accurate reconstruction of near-epipolar line segments from stereo aerial images

Research output: Contribution to journalArticleResearchpeer review

Authors

  • Ali Özgün Ok
  • Jan Dirk Wegner
  • Christian Heipke
  • Franz Rottensteiner
  • Uwe Sörgel
  • Vedat Toprak

External Research Organisations

  • Mersin University
  • Orta Dogu Technical University
View graph of relations

Details

Original languageEnglish
Pages (from-to)345-358
Number of pages14
JournalPhotogrammetrie, Fernerkundung, Geoinformation
Volume2012
Issue number4
Publication statusPublished - 1 Aug 2012

Abstract

In this study, a new 3D reconstructionapproach for line segments from stereo images isproposed which covers the case that the image linesare nearly-aligned (=10o) with the epipolar line.The method manipulates the redundancy inherentin line pair-relations to generate artificial 3D pointsand utilizes those entities during the estimationprocess to improve the reconstruction of the linesegments. The best points for the reconstruction areselected based on a newly proposed weight function.To test the performance of the proposed approach,we selected three test patches over a builtuparea of the city of Vaihingen, Germany. Theproposed approach produced a highly promisingreconstruction performance for line segments thatare nearly aligned with the epipolar line.

Keywords

    Line matching, Line reconstruction, Near epipolar line segments, Projectivegeometry, Stereo aerial images

ASJC Scopus subject areas

Cite this

Accurate reconstruction of near-epipolar line segments from stereo aerial images. / Ok, Ali Özgün; Wegner, Jan Dirk; Heipke, Christian et al.
In: Photogrammetrie, Fernerkundung, Geoinformation, Vol. 2012, No. 4, 01.08.2012, p. 345-358.

Research output: Contribution to journalArticleResearchpeer review

Ok, AÖ, Wegner, JD, Heipke, C, Rottensteiner, F, Sörgel, U & Toprak, V 2012, 'Accurate reconstruction of near-epipolar line segments from stereo aerial images', Photogrammetrie, Fernerkundung, Geoinformation, vol. 2012, no. 4, pp. 345-358. https://doi.org/10.1127/1432-8364/2012/0122
Ok, A. Ö., Wegner, J. D., Heipke, C., Rottensteiner, F., Sörgel, U., & Toprak, V. (2012). Accurate reconstruction of near-epipolar line segments from stereo aerial images. Photogrammetrie, Fernerkundung, Geoinformation, 2012(4), 345-358. https://doi.org/10.1127/1432-8364/2012/0122
Ok AÖ, Wegner JD, Heipke C, Rottensteiner F, Sörgel U, Toprak V. Accurate reconstruction of near-epipolar line segments from stereo aerial images. Photogrammetrie, Fernerkundung, Geoinformation. 2012 Aug 1;2012(4):345-358. doi: 10.1127/1432-8364/2012/0122
Ok, Ali Özgün ; Wegner, Jan Dirk ; Heipke, Christian et al. / Accurate reconstruction of near-epipolar line segments from stereo aerial images. In: Photogrammetrie, Fernerkundung, Geoinformation. 2012 ; Vol. 2012, No. 4. pp. 345-358.
Download
@article{02dd6ebb6be64ca08830ca55db848798,
title = "Accurate reconstruction of near-epipolar line segments from stereo aerial images",
abstract = "In this study, a new 3D reconstructionapproach for line segments from stereo images isproposed which covers the case that the image linesare nearly-aligned (=10o) with the epipolar line.The method manipulates the redundancy inherentin line pair-relations to generate artificial 3D pointsand utilizes those entities during the estimationprocess to improve the reconstruction of the linesegments. The best points for the reconstruction areselected based on a newly proposed weight function.To test the performance of the proposed approach,we selected three test patches over a builtuparea of the city of Vaihingen, Germany. Theproposed approach produced a highly promisingreconstruction performance for line segments thatare nearly aligned with the epipolar line.",
keywords = "Line matching, Line reconstruction, Near epipolar line segments, Projectivegeometry, Stereo aerial images",
author = "Ok, {Ali {\"O}zg{\"u}n} and Wegner, {Jan Dirk} and Christian Heipke and Franz Rottensteiner and Uwe S{\"o}rgel and Vedat Toprak",
year = "2012",
month = aug,
day = "1",
doi = "10.1127/1432-8364/2012/0122",
language = "English",
volume = "2012",
pages = "345--358",
journal = "Photogrammetrie, Fernerkundung, Geoinformation",
issn = "1432-8364",
publisher = "E. Schweizerbartsche Verlagsbuchhandlung",
number = "4",

}

Download

TY - JOUR

T1 - Accurate reconstruction of near-epipolar line segments from stereo aerial images

AU - Ok, Ali Özgün

AU - Wegner, Jan Dirk

AU - Heipke, Christian

AU - Rottensteiner, Franz

AU - Sörgel, Uwe

AU - Toprak, Vedat

PY - 2012/8/1

Y1 - 2012/8/1

N2 - In this study, a new 3D reconstructionapproach for line segments from stereo images isproposed which covers the case that the image linesare nearly-aligned (=10o) with the epipolar line.The method manipulates the redundancy inherentin line pair-relations to generate artificial 3D pointsand utilizes those entities during the estimationprocess to improve the reconstruction of the linesegments. The best points for the reconstruction areselected based on a newly proposed weight function.To test the performance of the proposed approach,we selected three test patches over a builtuparea of the city of Vaihingen, Germany. Theproposed approach produced a highly promisingreconstruction performance for line segments thatare nearly aligned with the epipolar line.

AB - In this study, a new 3D reconstructionapproach for line segments from stereo images isproposed which covers the case that the image linesare nearly-aligned (=10o) with the epipolar line.The method manipulates the redundancy inherentin line pair-relations to generate artificial 3D pointsand utilizes those entities during the estimationprocess to improve the reconstruction of the linesegments. The best points for the reconstruction areselected based on a newly proposed weight function.To test the performance of the proposed approach,we selected three test patches over a builtuparea of the city of Vaihingen, Germany. Theproposed approach produced a highly promisingreconstruction performance for line segments thatare nearly aligned with the epipolar line.

KW - Line matching

KW - Line reconstruction

KW - Near epipolar line segments

KW - Projectivegeometry

KW - Stereo aerial images

UR - http://www.scopus.com/inward/record.url?scp=84865202754&partnerID=8YFLogxK

U2 - 10.1127/1432-8364/2012/0122

DO - 10.1127/1432-8364/2012/0122

M3 - Article

AN - SCOPUS:84865202754

VL - 2012

SP - 345

EP - 358

JO - Photogrammetrie, Fernerkundung, Geoinformation

JF - Photogrammetrie, Fernerkundung, Geoinformation

SN - 1432-8364

IS - 4

ER -