Occlusion ambiguities in motion

Davi Geiger; Kostas I. Diamantaras

doi:10.1007/3-540-57956-7_20

Occlusion ambiguities in motion

Davi Geiger, Kostas I. Diamantaras

Computer Science

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

Abstract

We address the problem of optical flow reconstruction and in particular the problem of resolving ambiguities near edges. They occur due to (i) the aperture problem and (ii) the occlusion problem, where pixels on both sides of an intensity edge are assigned the same velocity estimates (and confidence). However, these measurements are correct for just one side of the edge (the non occluded one). We note that the confidence measures are large at intensity edges and larger at the convex sides of the edges, i.e. inside corners, than at the concave side. We resolve the ambiguities through local interactions via coupled Markov random fields (MRF). The result is the detection of motion for regions of images with large global convexity.

Original language	English (US)
Title of host publication	Computer Vision — ECCV 1994 - 3rd European Conference on Computer Vision, Proceedings
Editors	Jan-Olof Eklundh
Publisher	Springer Verlag
Pages	175-180
Number of pages	6
ISBN (Print)	9783540579564
DOIs	https://doi.org/10.1007/3-540-57956-7_20
State	Published - 1994
Event	3rd European Conference on Computer Vision, ECCV 1994 - Stockholm, Sweden Duration: May 2 1994 → May 6 1994

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	800 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	3rd European Conference on Computer Vision, ECCV 1994
Country/Territory	Sweden
City	Stockholm
Period	5/2/94 → 5/6/94

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/3-540-57956-7_20

Cite this

Geiger, D., & Diamantaras, K. I. (1994). Occlusion ambiguities in motion. In J.-O. Eklundh (Ed.), Computer Vision — ECCV 1994 - 3rd European Conference on Computer Vision, Proceedings (pp. 175-180). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 800 LNCS). Springer Verlag. https://doi.org/10.1007/3-540-57956-7_20

Occlusion ambiguities in motion. / Geiger, Davi; Diamantaras, Kostas I.
Computer Vision — ECCV 1994 - 3rd European Conference on Computer Vision, Proceedings. ed. / Jan-Olof Eklundh. Springer Verlag, 1994. p. 175-180 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 800 LNCS).

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

Geiger, D & Diamantaras, KI 1994, Occlusion ambiguities in motion. in J-O Eklundh (ed.), Computer Vision — ECCV 1994 - 3rd European Conference on Computer Vision, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 800 LNCS, Springer Verlag, pp. 175-180, 3rd European Conference on Computer Vision, ECCV 1994, Stockholm, Sweden, 5/2/94. https://doi.org/10.1007/3-540-57956-7_20

Geiger D, Diamantaras KI. Occlusion ambiguities in motion. In Eklundh JO, editor, Computer Vision — ECCV 1994 - 3rd European Conference on Computer Vision, Proceedings. Springer Verlag. 1994. p. 175-180. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/3-540-57956-7_20

@inproceedings{249a469e06a4445782a932edbc594160,

title = "Occlusion ambiguities in motion",

abstract = "We address the problem of optical flow reconstruction and in particular the problem of resolving ambiguities near edges. They occur due to (i) the aperture problem and (ii) the occlusion problem, where pixels on both sides of an intensity edge are assigned the same velocity estimates (and confidence). However, these measurements are correct for just one side of the edge (the non occluded one). We note that the confidence measures are large at intensity edges and larger at the convex sides of the edges, i.e. inside corners, than at the concave side. We resolve the ambiguities through local interactions via coupled Markov random fields (MRF). The result is the detection of motion for regions of images with large global convexity.",

author = "Davi Geiger and Diamantaras, {Kostas I.}",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 1994.; 3rd European Conference on Computer Vision, ECCV 1994 ; Conference date: 02-05-1994 Through 06-05-1994",

year = "1994",

doi = "10.1007/3-540-57956-7_20",

language = "English (US)",

isbn = "9783540579564",

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

publisher = "Springer Verlag",

pages = "175--180",

editor = "Jan-Olof Eklundh",

booktitle = "Computer Vision — ECCV 1994 - 3rd European Conference on Computer Vision, Proceedings",

}

TY - GEN

T1 - Occlusion ambiguities in motion

AU - Geiger, Davi

AU - Diamantaras, Kostas I.

N1 - Publisher Copyright: © Springer-Verlag Berlin Heidelberg 1994.

PY - 1994

Y1 - 1994

N2 - We address the problem of optical flow reconstruction and in particular the problem of resolving ambiguities near edges. They occur due to (i) the aperture problem and (ii) the occlusion problem, where pixels on both sides of an intensity edge are assigned the same velocity estimates (and confidence). However, these measurements are correct for just one side of the edge (the non occluded one). We note that the confidence measures are large at intensity edges and larger at the convex sides of the edges, i.e. inside corners, than at the concave side. We resolve the ambiguities through local interactions via coupled Markov random fields (MRF). The result is the detection of motion for regions of images with large global convexity.

AB - We address the problem of optical flow reconstruction and in particular the problem of resolving ambiguities near edges. They occur due to (i) the aperture problem and (ii) the occlusion problem, where pixels on both sides of an intensity edge are assigned the same velocity estimates (and confidence). However, these measurements are correct for just one side of the edge (the non occluded one). We note that the confidence measures are large at intensity edges and larger at the convex sides of the edges, i.e. inside corners, than at the concave side. We resolve the ambiguities through local interactions via coupled Markov random fields (MRF). The result is the detection of motion for regions of images with large global convexity.

UR - http://www.scopus.com/inward/record.url?scp=85027592012&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85027592012&partnerID=8YFLogxK

U2 - 10.1007/3-540-57956-7_20

DO - 10.1007/3-540-57956-7_20

M3 - Conference contribution

AN - SCOPUS:85027592012

SN - 9783540579564

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

SP - 175

EP - 180

BT - Computer Vision — ECCV 1994 - 3rd European Conference on Computer Vision, Proceedings

A2 - Eklundh, Jan-Olof

PB - Springer Verlag

T2 - 3rd European Conference on Computer Vision, ECCV 1994

Y2 - 2 May 1994 through 6 May 1994

ER -

Occlusion ambiguities in motion

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this