Restauración de Imágenes en Escala de Grises usando una red Neuronal Polinómica Discreta Recurrente

Abstract

In this work, a recurrent discrete polynomial neural network is constructed from the fixed-point attractors of cubic polynomials; and is used in grayscale image restoration. In addition, a criterion is provided for the assignment of values to the synaptic weights of the neural network, which will guarantee the stability of the neural network.