Projective reconstruction from incomplete trajectories by global and local constraints

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Authors

View graph of relations

Details

Original languageEnglish
Title of host publicationProceedings
Subtitle of host publication2011 Conference for Visual Media Production, CVMP 2011
Pages77-86
Number of pages10
Publication statusPublished - 2011
Event8th European Conference on Visual Media Production, CVMP 2011 - London, United Kingdom (UK)
Duration: 16 Nov 201117 Nov 2011

Publication series

NameProceedings - 2011 Conference for Visual Media Production, CVMP 2011

Abstract

The paper deals with projective shape and motion reconstruction by subspace iterations. A prerequisite of factorization-style algorithms is that all feature points need be observed in all images, a condition which is hardly realistic in real videos. We therefore address the problem of estimating structure and motion considering missing features. The proposed algorithm does not require initialization and uniformly handles all available data. The computed solution is global in the sense that it does not merge partial solutions incrementally or hierarchically. The global cost due to the factorization is further amended by local constraints to regularize and stabilize the estimations. It is shown how both costs can be jointly minimized in the presence of unobserved points. By synthetic and real image sequences with up to 60% missing data we demonstrate that our algorithm is accurate and reliable.

Keywords

    factorization, projective 3d-reconstruction

ASJC Scopus subject areas

Cite this

Projective reconstruction from incomplete trajectories by global and local constraints. / Ackermann, Hanno; Rosenhahn, Bodo.
Proceedings: 2011 Conference for Visual Media Production, CVMP 2011. 2011. p. 77-86 6103278 (Proceedings - 2011 Conference for Visual Media Production, CVMP 2011).

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Ackermann, H & Rosenhahn, B 2011, Projective reconstruction from incomplete trajectories by global and local constraints. in Proceedings: 2011 Conference for Visual Media Production, CVMP 2011., 6103278, Proceedings - 2011 Conference for Visual Media Production, CVMP 2011, pp. 77-86, 8th European Conference on Visual Media Production, CVMP 2011, London, United Kingdom (UK), 16 Nov 2011. https://doi.org/10.1109/CVMP.2011.15
Ackermann, H., & Rosenhahn, B. (2011). Projective reconstruction from incomplete trajectories by global and local constraints. In Proceedings: 2011 Conference for Visual Media Production, CVMP 2011 (pp. 77-86). Article 6103278 (Proceedings - 2011 Conference for Visual Media Production, CVMP 2011). https://doi.org/10.1109/CVMP.2011.15
Ackermann H, Rosenhahn B. Projective reconstruction from incomplete trajectories by global and local constraints. In Proceedings: 2011 Conference for Visual Media Production, CVMP 2011. 2011. p. 77-86. 6103278. (Proceedings - 2011 Conference for Visual Media Production, CVMP 2011). doi: 10.1109/CVMP.2011.15
Ackermann, Hanno ; Rosenhahn, Bodo. / Projective reconstruction from incomplete trajectories by global and local constraints. Proceedings: 2011 Conference for Visual Media Production, CVMP 2011. 2011. pp. 77-86 (Proceedings - 2011 Conference for Visual Media Production, CVMP 2011).
Download
@inproceedings{6a36fb1c339545b2b04307bd03db0741,
title = "Projective reconstruction from incomplete trajectories by global and local constraints",
abstract = "The paper deals with projective shape and motion reconstruction by subspace iterations. A prerequisite of factorization-style algorithms is that all feature points need be observed in all images, a condition which is hardly realistic in real videos. We therefore address the problem of estimating structure and motion considering missing features. The proposed algorithm does not require initialization and uniformly handles all available data. The computed solution is global in the sense that it does not merge partial solutions incrementally or hierarchically. The global cost due to the factorization is further amended by local constraints to regularize and stabilize the estimations. It is shown how both costs can be jointly minimized in the presence of unobserved points. By synthetic and real image sequences with up to 60% missing data we demonstrate that our algorithm is accurate and reliable.",
keywords = "factorization, projective 3d-reconstruction",
author = "Hanno Ackermann and Bodo Rosenhahn",
year = "2011",
doi = "10.1109/CVMP.2011.15",
language = "English",
isbn = "9780769546216",
series = "Proceedings - 2011 Conference for Visual Media Production, CVMP 2011",
pages = "77--86",
booktitle = "Proceedings",
note = "8th European Conference on Visual Media Production, CVMP 2011 ; Conference date: 16-11-2011 Through 17-11-2011",

}

Download

TY - GEN

T1 - Projective reconstruction from incomplete trajectories by global and local constraints

AU - Ackermann, Hanno

AU - Rosenhahn, Bodo

PY - 2011

Y1 - 2011

N2 - The paper deals with projective shape and motion reconstruction by subspace iterations. A prerequisite of factorization-style algorithms is that all feature points need be observed in all images, a condition which is hardly realistic in real videos. We therefore address the problem of estimating structure and motion considering missing features. The proposed algorithm does not require initialization and uniformly handles all available data. The computed solution is global in the sense that it does not merge partial solutions incrementally or hierarchically. The global cost due to the factorization is further amended by local constraints to regularize and stabilize the estimations. It is shown how both costs can be jointly minimized in the presence of unobserved points. By synthetic and real image sequences with up to 60% missing data we demonstrate that our algorithm is accurate and reliable.

AB - The paper deals with projective shape and motion reconstruction by subspace iterations. A prerequisite of factorization-style algorithms is that all feature points need be observed in all images, a condition which is hardly realistic in real videos. We therefore address the problem of estimating structure and motion considering missing features. The proposed algorithm does not require initialization and uniformly handles all available data. The computed solution is global in the sense that it does not merge partial solutions incrementally or hierarchically. The global cost due to the factorization is further amended by local constraints to regularize and stabilize the estimations. It is shown how both costs can be jointly minimized in the presence of unobserved points. By synthetic and real image sequences with up to 60% missing data we demonstrate that our algorithm is accurate and reliable.

KW - factorization

KW - projective 3d-reconstruction

UR - http://www.scopus.com/inward/record.url?scp=84855841716&partnerID=8YFLogxK

U2 - 10.1109/CVMP.2011.15

DO - 10.1109/CVMP.2011.15

M3 - Conference contribution

AN - SCOPUS:84855841716

SN - 9780769546216

T3 - Proceedings - 2011 Conference for Visual Media Production, CVMP 2011

SP - 77

EP - 86

BT - Proceedings

T2 - 8th European Conference on Visual Media Production, CVMP 2011

Y2 - 16 November 2011 through 17 November 2011

ER -

By the same author(s)