Please use this identifier to cite or link to this item: http://hdl.handle.net/10553/53228
Title: Newton filters: a new class of neuron-like discrete filters and an application to image processing
Authors: Quesada-Arencibia, A 
Aleman-Flores, M 
Moreno-Diaz, R 
UNESCO Clasification: 120304 Inteligencia artificial
220990 Tratamiento digital. Imágenes
Issue Date: 1999
Journal: Artificial Neural Nets And Genetic Algorithms
Conference: International Conference on Artificial Neural Networks and Genetic Algorithms (ICANNGA 99) 
Abstract: The functions of dentritic trees of neurons have always attracted neuroscientists. Dendrites are extensions of the soma that allows the cell to increase the area where it receives information from. The number of dentritic ramifications is not constant, it depends on the neuron and varies from one to the other. Moreover, each dentritic tree can be subdivided in a. complex form leading to a characteristic tree structure[1]. Two of their immediately-related properties regarding information processing (e.g. convergence and divergence of lines) have originated considerable research on completeness of computation and reliability of transmission [2]. In this paper a layered structure consisting in the interconnections of a set of simple functional units and inspired in the dentritic connections of real neurons, is suggested and in a parallel way we analyze the computing properties and characteristics together with a possible application in Image Processing. The interest of this structure lies in its similarity with the structure of the retinae receptive fields and even with dentritic trees of retinal neurons.
URI: http://hdl.handle.net/10553/53228
ISBN: 3-211-83364-1
Source: Artificial Neural Nets And Genetic Algorithms, p. 28-34
Appears in Collections:Actas de congresos
Show full item record

Page view(s)

36
checked on May 3, 2021

Google ScholarTM

Check

Altmetric


Share



Export metadata



Items in accedaCRIS are protected by copyright, with all rights reserved, unless otherwise indicated.