Please use this identifier to cite or link to this item:
http://hdl.handle.net/2445/33830
Title: | Line search multilevel optimization as computational methods for dense optical flow |
Author: | Kalmoun, El Mostafa Garrido Ostermann, Lluís Caselles, Vicente |
Keywords: | Informàtica Equacions diferencials Programació (Matemàtica) Càlcul de variacions Computer science Differential equations Mathematical programming Calculus of variations |
Issue Date: | 23-Jun-2011 |
Publisher: | Society for Industrial and Applied Mathematics |
Abstract: | We evaluate the performance of different optimization techniques developed in the context of optical flow computation with different variational models. In particular, based on truncated Newton methods (TN) that have been an effective approach for large-scale unconstrained optimization, we de- velop the use of efficient multilevel schemes for computing the optical flow. More precisely, we evaluate the performance of a standard unidirectional mul- tilevel algorithm - called multiresolution optimization (MR/OPT), to a bidrec- tional multilevel algorithm - called full multigrid optimization (FMG/OPT). The FMG/OPT algorithm treats the coarse grid correction as an optimiza- tion search direction and eventually scales it using a line search. Experimental results on different image sequences using four models of optical flow com- putation show that the FMG/OPT algorithm outperforms both the TN and MR/OPT algorithms in terms of the computational work and the quality of the optical flow estimation. |
Note: | Reproducció del document publicat a: http://dx.doi.org/10.1137/100807405 |
It is part of: | SIAM Journal On Imaging Sciences, 2011, vol. 4, num. 2, p. 695-722 |
URI: | http://hdl.handle.net/2445/33830 |
Related resource: | http://dx.doi.org/10.1137/100807405 |
ISSN: | 1936-4954 |
Appears in Collections: | Articles publicats en revistes (Matemàtiques i Informàtica) |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
603330.pdf | 645.73 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.