We describe an application of 1-Dimensional Kohonen Networks to the classification of color 2D images which has been evaluated in Popocatepetl Volcano's images. The Popocatepetl, located in the limits of the State of Puebla in Mexico, is active and under monitoring since 1997. We address the question if our application of the Kohonen Network classifies according to the total intensity color of an image or, if it classifies according to the connectivity, i.e. the topology, between the pixels that compose an image. To support the hypothesis that our system classifies according to the topology of the pixels in the images, we will present two approaches based a) in the evaluation of the classification given by the network when the pixels in the images are permuted; and, b) when an additional metric to the Euclidean distance is introduced. Index Terms - 1-Dimensional Kohonen Networks, Automated Image Classification, Metrics on Euclidean Spaces, Non-Supervised Classification.
[1]
Fernando Antonio Aguilera Ramírez,et al.
Non-supervised Classification of 2D Color Images Using Kohonen Networks and a Novel Metric
,
2005,
CIARP.
[2]
Ricardo Pérez-Aguila,et al.
A procedure for comparing color 2-dimensional images through their extrusions to the 5-dimensional colorspace
,
2005,
15th International Conference on Electronics, Communications and Computers (CONIELECOMP'05).
[3]
Helge J. Ritter,et al.
Neural computation and self-organizing maps - an introduction
,
1992,
Computation and neural systems series.
[4]
Francisco Casacuberta Nolla.
Redes Neuronales Artificiales
,
1998
.
[5]
D. Signorini,et al.
Neural networks
,
1995,
The Lancet.
[6]
Josiane Zerubia,et al.
Bayesian image classification using Markov random fields
,
1996,
Image Vis. Comput..