Optical flow estimation using Laplacian Mesh Energy

Wenbin Li, Darren Cosker, Matthew Brown, Rui Tang

Research output: Contribution to conferencePaper

16 Citations (Scopus)

Abstract

In this paper we present a novel non-rigid optical flow algorithm for dense image correspondence and non-rigid registration. The algorithm uses a unique Laplacian Mesh Energy term to encourage local smoothness whilst simultaneously preserving non-rigid deformation. Laplacian deformation approaches have become popular in graphics research as they enable mesh deformations to preserve local surface shape. In this work we propose a novel Laplacian Mesh Energy formula to ensure such sensible local deformations between image pairs. We express this wholly within the optical flow optimization, and show its application in a novel coarse-to-fine pyramidal approach. Our algorithm achieves the state-of-the-art performance in all trials on the Garg et al. dataset, and top tier performance on the Middlebury evaluation.
Original languageEnglish
Pages2435-2442
Number of pages8
DOIs
Publication statusPublished - 30 Jun 2013
EventIEEE International Conference on Computer Vision and Pattern Recognition (CVPR) - Oregon, US, UK United Kingdom
Duration: 25 Jun 201327 Jun 2013

Conference

ConferenceIEEE International Conference on Computer Vision and Pattern Recognition (CVPR)
CountryUK United Kingdom
CityOregon, US
Period25/06/1327/06/13

Fingerprint

Optical flows

Cite this

Li, W., Cosker, D., Brown, M., & Tang, R. (2013). Optical flow estimation using Laplacian Mesh Energy. 2435-2442. Paper presented at IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Oregon, US, UK United Kingdom. https://doi.org/10.1109/CVPR.2013.315

Optical flow estimation using Laplacian Mesh Energy. / Li, Wenbin; Cosker, Darren; Brown, Matthew; Tang, Rui.

2013. 2435-2442 Paper presented at IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Oregon, US, UK United Kingdom.

Research output: Contribution to conferencePaper

Li, W, Cosker, D, Brown, M & Tang, R 2013, 'Optical flow estimation using Laplacian Mesh Energy' Paper presented at IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Oregon, US, UK United Kingdom, 25/06/13 - 27/06/13, pp. 2435-2442. https://doi.org/10.1109/CVPR.2013.315
Li W, Cosker D, Brown M, Tang R. Optical flow estimation using Laplacian Mesh Energy. 2013. Paper presented at IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Oregon, US, UK United Kingdom. https://doi.org/10.1109/CVPR.2013.315
Li, Wenbin ; Cosker, Darren ; Brown, Matthew ; Tang, Rui. / Optical flow estimation using Laplacian Mesh Energy. Paper presented at IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), Oregon, US, UK United Kingdom.8 p.
@conference{c3687303f5ab41a2995f75fd3698d016,
title = "Optical flow estimation using Laplacian Mesh Energy",
abstract = "In this paper we present a novel non-rigid optical flow algorithm for dense image correspondence and non-rigid registration. The algorithm uses a unique Laplacian Mesh Energy term to encourage local smoothness whilst simultaneously preserving non-rigid deformation. Laplacian deformation approaches have become popular in graphics research as they enable mesh deformations to preserve local surface shape. In this work we propose a novel Laplacian Mesh Energy formula to ensure such sensible local deformations between image pairs. We express this wholly within the optical flow optimization, and show its application in a novel coarse-to-fine pyramidal approach. Our algorithm achieves the state-of-the-art performance in all trials on the Garg et al. dataset, and top tier performance on the Middlebury evaluation.",
author = "Wenbin Li and Darren Cosker and Matthew Brown and Rui Tang",
year = "2013",
month = "6",
day = "30",
doi = "10.1109/CVPR.2013.315",
language = "English",
pages = "2435--2442",
note = "IEEE International Conference on Computer Vision and Pattern Recognition (CVPR) ; Conference date: 25-06-2013 Through 27-06-2013",

}

TY - CONF

T1 - Optical flow estimation using Laplacian Mesh Energy

AU - Li, Wenbin

AU - Cosker, Darren

AU - Brown, Matthew

AU - Tang, Rui

PY - 2013/6/30

Y1 - 2013/6/30

N2 - In this paper we present a novel non-rigid optical flow algorithm for dense image correspondence and non-rigid registration. The algorithm uses a unique Laplacian Mesh Energy term to encourage local smoothness whilst simultaneously preserving non-rigid deformation. Laplacian deformation approaches have become popular in graphics research as they enable mesh deformations to preserve local surface shape. In this work we propose a novel Laplacian Mesh Energy formula to ensure such sensible local deformations between image pairs. We express this wholly within the optical flow optimization, and show its application in a novel coarse-to-fine pyramidal approach. Our algorithm achieves the state-of-the-art performance in all trials on the Garg et al. dataset, and top tier performance on the Middlebury evaluation.

AB - In this paper we present a novel non-rigid optical flow algorithm for dense image correspondence and non-rigid registration. The algorithm uses a unique Laplacian Mesh Energy term to encourage local smoothness whilst simultaneously preserving non-rigid deformation. Laplacian deformation approaches have become popular in graphics research as they enable mesh deformations to preserve local surface shape. In this work we propose a novel Laplacian Mesh Energy formula to ensure such sensible local deformations between image pairs. We express this wholly within the optical flow optimization, and show its application in a novel coarse-to-fine pyramidal approach. Our algorithm achieves the state-of-the-art performance in all trials on the Garg et al. dataset, and top tier performance on the Middlebury evaluation.

UR - http://dx.doi.org/10.1109/CVPR.2013.315

UR - http://www.cv-foundation.org/openaccess/CVPR2013.py

U2 - 10.1109/CVPR.2013.315

DO - 10.1109/CVPR.2013.315

M3 - Paper

SP - 2435

EP - 2442

ER -