Segmentation d'images en couleur par classification morphologique non supervisée

Authors

Thierry Géraud, Pierre-Yves Strub, Jérôme Darbon

Where

Proceedings of the International Conference on Image and Signal Processing (ICISP)

Place

Agadir, Morocco

Type

inproceedings

Publisher

Faculty of Sciences at Ibn Zohr University, Morocco

Keywords

Image

Date

2001-05-01

Abstract

In this paper, we present an original method to segment color images using a classification of the image histogram in the 3D color space. As color modes in natural images usually do not fit a well-known statistical model, we propose a classifier that rely on mathematical morphology and, more particularly, on the watershed algorithm. We show on various images that the expected color modes are correctly identified and, in order to obtain coherent region, we extend the method to make the segmentation contextual.

Bibtex (lrde.bib)

@InProceedings{	  geraud.01.icisp,
  author	= {Thierry G\'eraud and Pierre-Yves Strub and J\'er\^ome
		  Darbon},
  title		= {Segmentation d'images en couleur par classification
		  morphologique non supervis\'ee},
  booktitle	= {Proceedings of the International Conference on Image and
		  Signal Processing (ICISP)},
  year		= 2001,
  pages		= {387--394},
  address	= {Agadir, Morocco},
  month		= may,
  publisher	= {Faculty of Sciences at Ibn Zohr University, Morocco},
  note		= {In French},
  project	= {Image},
  abstract	= {In this paper, we present an original method to segment
		  color images using a classification of the image histogram
		  in the 3D color space. As color modes in natural images
		  usually do not fit a well-known statistical model, we
		  propose a classifier that rely on mathematical morphology
		  and, more particularly, on the watershed algorithm. We show
		  on various images that the expected color modes are
		  correctly identified and, in order to obtain coherent
		  region, we extend the method to make the segmentation
		  contextual.}
}