CNN BASED ON MULTI-VALUED NEURON AS A MODEL
OF ASSOCIATIVE MEMORY FOR GREY-SCALE IMAGES
Naum N.Aizenberg', Igor N.Aizenberg"
University of Uzhgorod, professor of the Department of Cybernetics;
Geroev Stalingrada 28, kv.49, Uzhgorod, 249015. Ukraine
Phone: (+7 03122) 23908, Fax: (+7 03122) 36120
Joint Venture PGD, Research Department Chief, Dr.Sc.;
Engelsa 27, kv.32, Uzhgorod, 294015, Ukraine
Phone: (+703122) 63269, Fax: (+703122) 36120
This work is supported by Maltian company DKL Ltd -
198 Old Bakery Street, Valetta, Malta. Fax: (+356)221893
In this paper we consider mathemntical model of multi-valued neural element
which is suggested to be used as one of the alternatives of the basic CNN
element. On the basis of such CNN we offer the model of the associative
memory for storing grey-scale images. Efficient learning algorithm for
neural element and associative memory is presented in the paper.