Motion detail preserving optical flow estimation

Li Xu; Jiaya Jia; Yasuyuki Matsushita

doi:10.1109/CVPR.2010.5539820

Motion detail preserving optical flow estimation

Li Xu^*, Jiaya Jia, Yasuyuki Matsushita

^*Corresponding author for this work

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

84 Citations (Scopus)

Abstract

We discuss the cause of a severe optical flow estimation problem that fine motion structures cannot always be correctly reconstructed in the commonly employed multi-scale variational framework. Our major finding is that significant and abrupt displacement transition wrecks small-scale motion structures in the coarse-to-fine refinement. A novel optical flow estimation method is proposed in this paper to address this issue, which reduces the reliance of the flow estimates on their initial values propagated from the coarser level and enables recovering many motion details in each scale. The contribution of this paper also includes adaption of the objective function and development of a new optimization procedure. The effectiveness of our method is borne out by experiments for both large- and small-displacement optical flow estimation.

Original language	English
Title of host publication	2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010
Pages	1293-1300
Number of pages	8
DOIs	https://doi.org/10.1109/CVPR.2010.5539820
Publication status	Published - 2010
Externally published	Yes
Event	2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010 - San Francisco, CA, United States Duration: 13 Jun 2010 → 18 Jun 2010

Publication series

Name	Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
ISSN (Print)	1063-6919

Conference

Conference	2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010
Country/Territory	United States
City	San Francisco, CA
Period	13/06/10 → 18/06/10

Access to Document

10.1109/CVPR.2010.5539820

Cite this

@inproceedings{2ddadb9f13ba412a8f317c7aa3006447,

title = "Motion detail preserving optical flow estimation",

abstract = "We discuss the cause of a severe optical flow estimation problem that fine motion structures cannot always be correctly reconstructed in the commonly employed multi-scale variational framework. Our major finding is that significant and abrupt displacement transition wrecks small-scale motion structures in the coarse-to-fine refinement. A novel optical flow estimation method is proposed in this paper to address this issue, which reduces the reliance of the flow estimates on their initial values propagated from the coarser level and enables recovering many motion details in each scale. The contribution of this paper also includes adaption of the objective function and development of a new optimization procedure. The effectiveness of our method is borne out by experiments for both large- and small-displacement optical flow estimation.",

author = "Li Xu and Jiaya Jia and Yasuyuki Matsushita",

year = "2010",

doi = "10.1109/CVPR.2010.5539820",

language = "English",

isbn = "9781424469840",

series = "Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition",

pages = "1293--1300",

booktitle = "2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010",

note = "2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010 ; Conference date: 13-06-2010 Through 18-06-2010",

}

Xu, L, Jia, J & Matsushita, Y 2010, Motion detail preserving optical flow estimation. in 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010., 5539820, Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 1293-1300, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010, San Francisco, CA, United States, 13/06/10. https://doi.org/10.1109/CVPR.2010.5539820

Motion detail preserving optical flow estimation. / Xu, Li; Jia, Jiaya; Matsushita, Yasuyuki.
2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010. 2010. p. 1293-1300 5539820 (Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition).

Research output: Chapter in Book/Conference Proceeding/Report › Conference Paper published in a book › peer-review

TY - GEN

T1 - Motion detail preserving optical flow estimation

AU - Xu, Li

AU - Jia, Jiaya

AU - Matsushita, Yasuyuki

PY - 2010

Y1 - 2010

N2 - We discuss the cause of a severe optical flow estimation problem that fine motion structures cannot always be correctly reconstructed in the commonly employed multi-scale variational framework. Our major finding is that significant and abrupt displacement transition wrecks small-scale motion structures in the coarse-to-fine refinement. A novel optical flow estimation method is proposed in this paper to address this issue, which reduces the reliance of the flow estimates on their initial values propagated from the coarser level and enables recovering many motion details in each scale. The contribution of this paper also includes adaption of the objective function and development of a new optimization procedure. The effectiveness of our method is borne out by experiments for both large- and small-displacement optical flow estimation.

AB - We discuss the cause of a severe optical flow estimation problem that fine motion structures cannot always be correctly reconstructed in the commonly employed multi-scale variational framework. Our major finding is that significant and abrupt displacement transition wrecks small-scale motion structures in the coarse-to-fine refinement. A novel optical flow estimation method is proposed in this paper to address this issue, which reduces the reliance of the flow estimates on their initial values propagated from the coarser level and enables recovering many motion details in each scale. The contribution of this paper also includes adaption of the objective function and development of a new optimization procedure. The effectiveness of our method is borne out by experiments for both large- and small-displacement optical flow estimation.

UR - https://www.webofscience.com/wos/woscc/full-record/WOS:000287417501043

UR - https://openalex.org/W4242007601

UR - https://www.scopus.com/pages/publications/77955995204

U2 - 10.1109/CVPR.2010.5539820

DO - 10.1109/CVPR.2010.5539820

M3 - Conference Paper published in a book

SN - 9781424469840

T3 - Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition

SP - 1293

EP - 1300

BT - 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010

T2 - 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010

Y2 - 13 June 2010 through 18 June 2010

ER -

Motion detail preserving optical flow estimation

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this