Research and realization of digital recognition based on hopfield neural network

Abstract

Digital recognition is an important aspect of computer model recognition. It has a very good prospect, as well as basis for post-processing. Discrete hopfield neural network simulates memory mechanism of biological neural network. Specifically, it first learns memory samples, then associates original figure according to noise figure to be identified. The paper identifys figure which have sufferd noise pollution with the use of discrete hopfield neural network. Finally, the issue puts forward certain important proposals. © 2012 Springer-Verlag Berlin Heidelberg.