mk:iwann03

Summary

Morphological Clustering of the SOM for Multi-Dimensional Image Segmentation. Aureli Soria-Frisch and Mario Köppen. In Proc. Int. Workshop on Artificial Neural Networks, IWANN, Mao, Balearic Islands,, pages 582-588, 2003. (URL)

Abstract

New imaging sensors and technologies challenge the application of traditional computer vision methodologies due to an increment of the image dimensionality. Images processed in different application fields are not any more restricted to the grayvalue image domain, but take into consideration images of larger dimensionality. Color, multisensorial and satellite images are some examples being used in different application fields. This increment in the dimensionality of the problems related to the utilization of these larger feature spaces has been already characterized as a problem denoted by the ``curse of dimensionality'' in pattern recognition. Taking into consideration feature spaces of large dimensions introduce some geometric anomalies, which hinder the interpretability of the results and thus the attainment of the expected ones [3]. The paper presents a hybrid framework, which makes use of a Self-Organizing Map (SOM) [2] and the fuzzy integral [1] in order to cope with the segmentation of images in high-dimensional feature spaces.

Bibtex entry

@INPROCEEDINGS { mk:iwann03,
    ABSTRACT = { New imaging sensors and technologies challenge the application of traditional computer vision methodologies due to an increment of the image dimensionality. Images processed in different application fields are not any more restricted to the grayvalue image domain, but take into consideration images of larger dimensionality. Color, multisensorial and satellite images are some examples being used in different application fields. This increment in the dimensionality of the problems related to the utilization of these larger feature spaces has been already characterized as a problem denoted by the ``curse of dimensionality'' in pattern recognition. Taking into consideration feature spaces of large dimensions introduce some geometric anomalies, which hinder the interpretability of the results and thus the attainment of the expected ones [3]. The paper presents a hybrid framework, which makes use of a Self-Organizing Map (SOM) [2] and the fuzzy integral [1] in order to cope with the segmentation of images in high-dimensional feature spaces. },
    AUTHOR = { Aureli Soria-Frisch and Mario Köppen },
    BOOKTITLE = { Proc. Int. Workshop on Artificial Neural Networks, IWANN, Mao, Balearic Islands, },
    MODIFIED = { 2008-02-28 14:52:33 +0900 },
    DOI = { 10.1007/3-540-44868-3_74 },
    HASABSTRACT = { Yes },
    PAGES = { 582--588 },
    PUBLISHER = { Springer-Verlag Heidelberg },
    SERIES = { LNCS 2686 },
    TITLE = { Morphological Clustering of the SOM for Multi-Dimensional Image Segmentation },
    URL = { http://www.springerlink.com/content/5ce0wwvk1kultxcr },
    YEAR = { 2003 },
    1 = { http://www.springerlink.com/content/5ce0wwvk1kultxcr },
    2 = { http://dx.doi.org/10.1007/3-540-44868-3_74 },
}

On small computer displays, you can hide this right bar by using the 'Hide' button above.

News

Next conferences COMPSAC 2014 (Vasteras, Sweden, July 2014), INCoS-2014 (Salerno, Italy, September 2014).

New edited book "Soft Computing in Industrial Applications", V. Snasel, P. Kroemer, M. Koeppen, G. Schaefer, Springer AISC 223, July 2013.