A hierarchical learning structure, combining a randomly-placed local window, a self-organising map and a local-voting scheme, has been developed for the unsupervised segmentation of textured images, which are modelled by Markov random fields. The system learns to progressively estimate model parameters, and hence classify the various textured regions. A globally correct segregation has consistently been obtained during extensive experiments on both synthetic and natural textured images.
References
-
-
1)
-
Lampinen, J., Oja, E.: `Self-organizing maps for spatial and temporal AR models', Proc. 6th Scandinavian Conf. on Image Analysis, 1989, p. 120–127.
-
2)
-
S. Lakshmanan ,
H. Derin
.
Simultaneous parameter estimation and segmentation of Gibbs random fieldsusing simulated annealing.
IEEE Trans.
,
8 ,
799 -
813
-
3)
-
P. Brodatz
.
(1966)
Textures: A photographic album for artists and designees.
-
4)
-
B.S. Manjunath ,
R. Chellappa
.
Unsupervised texture segmentation using Markov random field models.
IEEE Trans.
,
5 ,
478 -
482
-
5)
-
S. Geman ,
D. Geman
.
Stochastic relaxation, Gibbs distributions, and the Bayesian restorationof images.
IEEE Trans.
,
6 ,
721 -
741
-
6)
-
T. Kohonen
.
(1984)
Self-organization and associative memory.
-
7)
-
Yin, H., Allinson, N.M.: `Self-organised segmentation for textured images', Proc. ICANN'94, 1994, p. 1149–1152.
http://iet.metastore.ingenta.com/content/journals/10.1049/el_19941275
Related content
content/journals/10.1049/el_19941275
pub_keyword,iet_inspecKeyword,pub_concept
6
6