Computer Vision, Speech Communication &

Signal Processing Group

NTUA | ECE
Faculty | PhD Students | Collaborators
Journal | Book Chapters | Conference
Undergraduate | Graduate | Diploma Theses

Texture Analysis Using Modulation Features and Generative Models

Overview

This code contains the texture analysis functions for the paper `Texture Analysis and Segmentation Using Modulation Features, Generative Models, and Weighted Curve Evolution', by I. Kokkinos, G. Evangelopoulos and P. Maragos, appearing in IEEE Transactions on Pattern Analysis and Machine Intelligence, Volume 31, Issue 1, Jan. 2009 Page(s):142 - 157. See publications for this and related previous work.

This toolbox is developed in and written for MATLAB, with emphasis on efficient algorithm implementations for multiband image filtering, demodulation in amplitude (AM) and frequency (FM) signals via the regularized 2D discrete energy separation algorithm and probabilistic localization of texture, edge, smooth image regions. For a more detailed overview of the theory and the methods please refer to the presentation here. See also our texture related research web page here.

Author

Author of the toolbox is Iasonas Kokkinos, currently Assistant Professor at Ecole Centrale Paris, with partial contributions from Georgios Evangelopoulos.

Architecture

The provided functions include:

  • multi-scale & orientation filterbanks for gabors and edges
  • projection on the basis elements of the underlying generative models
  • demodulation with regularized/complex esa
  • channel selection based on the amplitude/teager/mdl criterion
  • texture/edge/smooth classification based on mdl criterion

Usage/Installation

To get demo results for 20 images:

  • open util/load_textures and modify the third line appropriately
  • type T00__batch_script on the matlab prompt

Parameters:

  • several parameters to play with are all accessible from the T00__batch_script
  • they are primarily related to the filterbank construction and the final classification stage

Downloads

For a preliminary release of the toolbox, along with basic documentation and function dependecies download the version here:

  • version 0.9 (preliminary release)[rar archive, 2MB, Jan. 28, 2008].
See the Release Notes for the change log.

A more complete version with how-to's, exntended documentation, examples and demos will appear here soon. For more information, please contact Iasonas Kokkinos by email at iasonas.kokkinos[at]ecp.fr.

Publications

  • I. Kokkinos, G. Evangelopoulos and P. Maragos,
    Texture Analysis and Segmentation using Modulation Features, Generative Models and Weighted Curve Evolution,
    IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 31, no. 1, pp. 142-157, Jan. 2009.
    [pdf] [bib]
  • G. Evangelopoulos, I. Kokkinos and P. Maragos,
    Advances in Variational Image Segmentation using AM-FM Models: Regularized Demodulation and Probabilistic Cue Integration,
    Proc. Int' l Workshop on Variational and Level Set Methods (VLSM-05), Beijing, China, Oct. 2005, Springer LNCS, vol. 3275, pp. 121-136.
    [pdf] [bib]
  • I. Kokkinos, G. Evangelopoulos and P. Maragos,
    Advances in Texture Analysis: Energy Dominant Component and Multiple Hypothesis Testing,
    Proc. IEEE Int' l Conf. on Image Processing (ICIP-04), Singapore, Oct. 2004, vol. 3, pp. 1509-1512.
    [pdf] [bib]
  • I. Kokkinos, G. Evangelopoulos and P. Maragos,
    Modulation-Feature Based Textured Image Segmentation Using Curve Evolution,
    Proc. IEEE Int' l Conf. on Image Processing (ICIP-04), Singapore, Oct. 2004, vol. 2, pp. 1204-1207.
    [pdf] [bib]

Presentations

  • I. Kokkinos,
    Texture Analysis and Segmentation using Modulation Models,
    Department of Mathematics, UCLA Image Processing Seminar.
    [slides]

Licensing

If you use this toolbox in your research, please cite the related journal paper.

Test images appearing in the paper are courtesy of the The Berkeley Segmentation Dataset. All are icluded in the toolbox for demo purposes.

Note:This is an image analysis and computer vision research software and thus comprehending the code requires familiarity with relevant ideas like image filtering. Also, due to time constraints, limited user support can be provided.

Acknowledgments

Financial support for this software has been provided by the research programs 'HRAKLEITOS', 'PENED-2001' of the Greek Secretariat for Research and Technology and the FP6 European Network of Excellence MUSCLE, under contract no. IST-FP6-507752. This is gratefully acknowledged.

Last Update: January 28, 2008
For comments or questions contact Iasonas Kokkinos.

Last modified: Monday, 02 February 2009 | Created by Nassos Katsamanis and George Papandreou