Dinámica neuronal para modelar el proceso de segmentación de estímulos visuales de color y textura
ISSN: 0214-932X
Year of publication: 1996
Volume: 29
Issue: 3
Pages: 16-27
Type: Article
More publications in: Informática y automática: revista de la Asociación Española de Informática y Automática
Abstract
In this paper we propose a dynamic neural network model for color image segmentation. The main objective has been to develop a model of the human visual system that combines both color and texture visual informations to obtain a coherent imagen segmentation. Imagen segmentation is achieved by transforming RGB signals to other color system that decorrelates chromatic and achromatic informations. This transformation is based on the opponent neurophysiological interactions in retinal neurons. The model consists of two modules called Color Oponent System and Segmentation Chromatic System, where the transformation and segmentation are achieved, respectively. Results obtained from the processing of real images supported the validity and applicability of the model to the solution of real problems in vision.