A combined method for texture analysis and its application

Yongping Zhang, Ruili Wang

Research output: Chapter in Book/Conference proceedingBook Chapterpeer-review

2 Citations (Scopus)

Abstract

In this paper, a rotational invariant feature set is introduced for texture classification, based on wavelet transformation in combination with co-occurrence probabilities. Using this combined method, through wavelet decomposition and reconstruction, an approximation image and a new details image are generated. Beside of using the statistic approximation and the new details respectively, the joint distribution of the original and the new detail image is computed, and seven novel digital features are derived from the joint probability. By combination with a MLP neural network, our method has successfully applied to pollen discrimination. In experiments with sixteen types of airborne pollen grains, more than 95 percent pollen images are correctly classified.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
EditorsMarian Bubak, Geert Dick van Albada, Peter M.A. Sloot, Jack J. Dongarra
PublisherSpringer Verlag
Pages413-416
Number of pages4
ISBN (Print)9783540221142
DOIs
Publication statusPublished - 2004
Externally publishedYes

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3036
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'A combined method for texture analysis and its application'. Together they form a unique fingerprint.

Cite this