Discovery - University of Dundee - Online Publications

Library & Learning Centre

Regular texture analysis as statistical model selection

Regular texture analysis as statistical model selection

Research output: Chapter in Book/Report/Conference proceedingConference contribution

View graph of relations

Authors

Research units

Info

Original languageEnglish
Title of host publicationComputer Vision – ECCV 2008
Subtitle of host publication10th European Conference on Computer Vision Marseille, France, October 12-18, 2008. Proceedings, Part IV
EditorsDavid Forsyth, Philip Torr, Andrew Zisserman
Place of PublicationBerlin
PublisherSpringer
Pages242-255
Number of pages14
ISBN (Electronic)9783540886938
ISBN (Print)9783540886921
DOIs
StatePublished - 2008
Event10th European Conference on Computer Vision - Marseille, France

Publication series

NameLecture notes in computer science
PublisherSpringer
Volume5305
ISSN (Print)0302-9743

Conference

Conference10th European Conference on Computer Vision
Abbreviated titleECCV 2008
CountryFrance
CityMarseille
Period12/10/0818/10/08
Internet address

Abstract

An approach to the analysis of images of regular texture is proposed in which lattice hypotheses are used to define statistical models. These models are then compared in terms of their ability to explain the image. A method based on this approach is described in which lattice hypotheses are generated using analysis of peaks in the image autocorrelation function, statistical models are based on Gaussian or Gaussian mixture clusters, and model comparison is performed using the marginal likelihood as approximated by the Bayes Information Criterion (BIC). Experiments on public domain regular texture images and a commercial textile image archive demonstrate substantially improved accuracy compared to two competing methods. The method is also used for classification of texture images as regular or irregular. An application to thumbnail image extraction is discussed.

Documents

DOI

Library & Learning Centre

Contact | Accessibility | Policy