A Spatially Correlated Mixture Model for Image Segmentation

Kosei KURISU; Nobuo SUEMATSU; Kazunori IWATA; Akira HAYASHI

doi:10.1587/transinf.2014EDP7307

A Spatially Correlated Mixture Model for Image Segmentation

Kosei KURISU, Nobuo SUEMATSU, Kazunori IWATA, Akira HAYASHI

Full Text Views

0

Cite this

Summary :

In image segmentation, finite mixture modeling has been widely used. In its simplest form, the spatial correlation among neighboring pixels is not taken into account, and its segmentation results can be largely deteriorated by noise in images. We propose a spatially correlated mixture model in which the mixing proportions of finite mixture models are governed by a set of underlying functions defined on the image space. The spatial correlation among pixels is introduced by putting a Gaussian process prior on the underlying functions. We can set the spatial correlation rather directly and flexibly by choosing the covariance function of the Gaussian process prior. The effectiveness of our model is demonstrated by experiments with synthetic and real images.

Publication: IEICE TRANSACTIONS on Information Vol.E98-D No.4 pp.930-937

Publication Date: 2015/04/01

Publicized: 2015/01/06

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2014EDP7307

Type of Manuscript: PAPER

Category: Image Recognition, Computer Vision

Authors

Kosei KURISU
  Hiroshima City University
Nobuo SUEMATSU
  Hiroshima City University
Kazunori IWATA
  Hiroshima City University
Akira HAYASHI
  Hiroshima City University

Keyword

image segmentation, Gaussian processes, mixture models

Cite this

Copy

Kosei KURISU, Nobuo SUEMATSU, Kazunori IWATA, Akira HAYASHI, "A Spatially Correlated Mixture Model for Image Segmentation" in IEICE TRANSACTIONS on Information, vol. E98-D, no. 4, pp. 930-937, April 2015, doi: 10.1587/transinf.2014EDP7307.
Abstract: In image segmentation, finite mixture modeling has been widely used. In its simplest form, the spatial correlation among neighboring pixels is not taken into account, and its segmentation results can be largely deteriorated by noise in images. We propose a spatially correlated mixture model in which the mixing proportions of finite mixture models are governed by a set of underlying functions defined on the image space. The spatial correlation among pixels is introduced by putting a Gaussian process prior on the underlying functions. We can set the spatial correlation rather directly and flexibly by choosing the covariance function of the Gaussian process prior. The effectiveness of our model is demonstrated by experiments with synthetic and real images.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2014EDP7307/_p

Copy

@ARTICLE{e98-d_4_930,
author={Kosei KURISU, Nobuo SUEMATSU, Kazunori IWATA, Akira HAYASHI, },
journal={IEICE TRANSACTIONS on Information},
title={A Spatially Correlated Mixture Model for Image Segmentation},
year={2015},
volume={E98-D},
number={4},
pages={930-937},
abstract={In image segmentation, finite mixture modeling has been widely used. In its simplest form, the spatial correlation among neighboring pixels is not taken into account, and its segmentation results can be largely deteriorated by noise in images. We propose a spatially correlated mixture model in which the mixing proportions of finite mixture models are governed by a set of underlying functions defined on the image space. The spatial correlation among pixels is introduced by putting a Gaussian process prior on the underlying functions. We can set the spatial correlation rather directly and flexibly by choosing the covariance function of the Gaussian process prior. The effectiveness of our model is demonstrated by experiments with synthetic and real images.},
keywords={},
doi={10.1587/transinf.2014EDP7307},
ISSN={1745-1361},
month={April},}

Copy

TY - JOUR
TI - A Spatially Correlated Mixture Model for Image Segmentation
T2 - IEICE TRANSACTIONS on Information
SP - 930
EP - 937
AU - Kosei KURISU
AU - Nobuo SUEMATSU
AU - Kazunori IWATA
AU - Akira HAYASHI
PY - 2015
DO - 10.1587/transinf.2014EDP7307
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E98-D
IS - 4
JA - IEICE TRANSACTIONS on Information
Y1 - April 2015
AB - In image segmentation, finite mixture modeling has been widely used. In its simplest form, the spatial correlation among neighboring pixels is not taken into account, and its segmentation results can be largely deteriorated by noise in images. We propose a spatially correlated mixture model in which the mixing proportions of finite mixture models are governed by a set of underlying functions defined on the image space. The spatial correlation among pixels is introduced by putting a Gaussian process prior on the underlying functions. We can set the spatial correlation rather directly and flexibly by choosing the covariance function of the Gaussian process prior. The effectiveness of our model is demonstrated by experiments with synthetic and real images.
ER -