Design of Neural Network System to Communicate a Blind Person with a Computer Using a Braille

Abstract

Artificial intelligence systems are widely used nowadays in solving different problems facing mankind. Artificial intelligence systems were developed based on how human brains can think and function; this distinguished character gives these systems the wide range of applications as compared to conventional systems. Among the artificial intelligent systems involves artificial neural intelligent, fuzzy systems and neural fuzzy systems.

Blind person interface to computer is among the problem which the world is facing. In this paper the Artificial Neural Network (ANN) is used in designing the Braille system which is capable of enabling the interface between a blind person and the computer. The Multilayer perceptron (MLP) with four layers and nineteen neural is used for the implementation of pattern recognition of the Braille. The pattern of the Braille is used as the inputs to the MLP whereby through MATLAB developed program the patterns training are achieved. The developed system is capable of enabling the blind person to typing letters, numbers and to enter different commands to the computer.