Presentation | 2010-10-23 Parametric control in a region-based coupled MRF model for coarse image region segmentation towards its VLSI implementation Haichao LIANG, Kenji MATSUZAKA, Kazuki NAKADA, Masato OKADA, Takashi MORIE, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | For visual image processing, coupled Markov random field (MRF) models have been proposed, and they are classified into boundary-based and region-based models, according to the interactions between the nodes corresponding to the image pixels. The original region-based MRF models have a local-minimum problem because of its discrete hidden variables, but an improved model has been proposed by introducing phase variables as continuous hidden variables to solve this problem. We have already modified this model by introducing piecewise linear (PWL) functions to simplify the model towards its VLSI implementation. In this paper, we investigate the controllability of the parameters of PWL functions, and evaluate the performance in the application of coarse image region segmentation. We have verified that the segmentation tuning and the processing speed can be improved by optimizing of the parameters. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | region-based coupled MRF model / boundary-based coupled MRF model / coarse image region segmentation / phase variables / piecewise linear functions / parametric control |
Paper # | NC2010-48 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2010/10/16(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Neurocomputing (NC) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Parametric control in a region-based coupled MRF model for coarse image region segmentation towards its VLSI implementation |
Sub Title (in English) | |
Keyword(1) | region-based coupled MRF model |
Keyword(2) | boundary-based coupled MRF model |
Keyword(3) | coarse image region segmentation |
Keyword(4) | phase variables |
Keyword(5) | piecewise linear functions |
Keyword(6) | parametric control |
1st Author's Name | Haichao LIANG |
1st Author's Affiliation | Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology() |
2nd Author's Name | Kenji MATSUZAKA |
2nd Author's Affiliation | Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology |
3rd Author's Name | Kazuki NAKADA |
3rd Author's Affiliation | Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology |
4th Author's Name | Masato OKADA |
4th Author's Affiliation | Graduate School of Frontier Sciences, The University of Tokyo |
5th Author's Name | Takashi MORIE |
5th Author's Affiliation | Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology |
Date | 2010-10-23 |
Paper # | NC2010-48 |
Volume (vol) | vol.110 |
Number (no) | 246 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |