Share Email Print

Proceedings Paper

Optimized texture classification by using hierarchical complex network measurements
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Texture characterization and classification remains an important issue in image processing and analysis. Much attention has been focused on methods involving spectral analysis and co-occurrence matrix, as well as more modern approaches such as those involving fractal dimension, entropy and criteria based in multiresolution. The present work addresses the problem of texture characterization in terms of complex networks: image pixels are represented as nodes and similarities between such pixels are mapped as links between the network nodes. It is verified that several types of textures present node degree distributions which are far distinct from those observed for random networks, suggesting complex organization of those textures. Traditional measurements of the network connectivity, including their respective hierarchical extensions, are then applied in order to obtain feature vectors from which the textures can be characterized and classified. The performance of such an approach is compared to co-occurrence methods, suggesting promising complementary perspectives.

Paper Details

Date Published: 9 February 2006
PDF: 8 pages
Proc. SPIE 6070, Machine Vision Applications in Industrial Inspection XIV, 60700Q (9 February 2006); doi: 10.1117/12.655592
Show Author Affiliations
T. Chalumeau, Univ. de Bourgogne, IUT Le Creusot, Le2i (France)
Univ. de São Paulo, IFSC (Brazil)
L. da F. Costa, Univ. de São Paulo, IFSC (Brazil)
O. Laligant, Univ. de Bourgogne, IUT Le Creusot, Le2i (France)
F. Meriaudeau, Univ. de Bourgogne, IUT Le Creusot, Le2i (France)

Published in SPIE Proceedings Vol. 6070:
Machine Vision Applications in Industrial Inspection XIV
Fabrice Meriaudeau; Kurt S. Niel, Editor(s)

© SPIE. Terms of Use
Back to Top