Please use this identifier to cite or link to this item:
Scopus Web of Science® Altmetric
Type: Conference paper
Title: Robust trajectory-space TV-L1 optical flow for non-rigid sequences
Author: Garg, R.
Roussos, A.
Agapito, L.
Citation: Lecture Notes in Artificial Intelligence, 2011 / Boykov, Y., Kahl, F., Lempitsky, V., Schmidt, F. (ed./s), vol.6819 LNCS, pp.300-314
Publisher: Springer
Issue Date: 2011
Series/Report no.: Lecture Notes in Computer Science; 6819
ISBN: 9783642230936
ISSN: 0302-9743
Conference Name: 8th International Conference on Energy Minimization Methods in Computer Vision and Pattern Recognition (EMMCVPR 2011) (25 Jul 2011 - 27 Jul 2011 : St. Petersburg, Russia)
Editor: Boykov, Y.
Kahl, F.
Lempitsky, V.
Schmidt, F.
Statement of
Ravi Garg, Anastasios Roussos, and Lourdes Agapito
Abstract: This paper deals with the problem of computing optical flow between each of the images in a sequence and a reference frame when the camera is viewing a non-rigid object. We exploit the high correlation between 2D trajectories of different points on the same non-rigid surface by assuming that the displacement sequence of any point can be expressed in a compact way as a linear combination of a low-rank motion basis. This subspace constraint effectively acts as a long term regularization leading to temporally consistent optical flow. We formulate it as a robust soft constraint within a variational framework by penalizing flow fields that lie outside the low-rank manifold. The resulting energy functional includes a quadratic relaxation term that allows to decouple the optimization of the brightness constancy and spatial regularization terms, leading to an efficient optimization scheme. We provide a new benchmark dataset, based on motion capture data of a flag waving in the wind, with dense ground truth optical flow for evaluation of multi-view optical flow of non-rigid surfaces. Our experiments, show that our proposed approach provides comparable or superior results to state of the art optical flow and dense non-rigid registration algorithms.
Rights: © Springer-Verlag Berlin Heidelberg 2011
DOI: 10.1007/978-3-642-23094-3_22
Published version:
Appears in Collections:Aurora harvest 3
Computer Science publications

Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.