Methods for recognizing multispectral images based on neural networks

This monograph presents the basic concepts methods and models of multispectral image recognition based on oscillatory and diagonalized neural networks in a logical sequence. In addition the monograph contains sections on the modified Hecht-Nielsen theorem optimizing the size of a neural network taking into account the nonlinear generalized error of the neural network model creating a pseudo-spin neural network with reduced learning time and creating a decision support model based on a diagonalized neural network.