Texture classification in the Wigner-Ville domain

Author

Dupont, F. ; Zhu, Y.M. ; Odet, C. ; Goutte, R.

Author_Institution

CREATIS, CNRS, Villeurbanne, France

Volume

1

fYear

1996

Firstpage

327

Abstract

A texture classification scheme is presented that is based on classifying spectral features derived from the two-dimensional Wigner-Ville distribution (WVD) using an iterative and unsupervised k-means classification technique. The proposed method is illustrated with the aid of both simulations and examples on physical microscopic images. The obtained results show that the WVD allows pertinent features of image regions to be extracted, and that the use of only these spectral features yields a satisfactory texture classification results

Keywords

Wigner distribution; feature extraction; image classification; image representation; image segmentation; image texture; iterative methods; spectral analysis; time-frequency analysis; Wigner-Ville domain; feature extraction; image regions; image representation; iterative classification technique; physical microscopic images; simulations; spectral features; texture classification; time-frequency tool; two-dimensional Wigner-Ville distribution; unsupervised k-means classification; Energy resolution; Image processing; Image texture analysis; Microscopy; Signal analysis; Spatial resolution; Spectral analysis; Spectrogram; Time frequency analysis; Two dimensional displays;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing, 1996., 3rd International Conference on

Conference_Location

Beijing

Print_ISBN

0-7803-2912-0

Type

conf

DOI

10.1109/ICSIGP.1996.567254

Filename

567254