Реалізаційна модель оптоелектронного навченого класифікатора

Abstract

In the master's qualification paper analyzes current trends related to methods and means of the model Optoelectronic trained classifier, the possible ways of encoding images (Arabic numerals) and analyzes possible element base for the implementation of e-classifier. An effective model representation of Arabic numerals, established base of reference samples and their classes. The block diagram of the e-classifier is created. Created algorithmic and software classification process. Tested program, where the object selected number "five" with varying degrees of noise.