Design and implementation of a robust system for recognizing alphabets using artifical neural network

Abstract

Language or pattern recognition is an ever-growing field in research. The motivation behind this work is the growing importance of foreign languages in everyday life. It is specifically focused on groups who might have to face a foreign language somewhere and can't seem to understand it. This will also be beneficial for partially visually impaired persons. It will help in recognizing the alphabets and letters written in a particular language stored in the database. The program was trained with different samples of a different letters of an alphabet and the data was stored. Using Artificial Neural Network and pattern recognition techniques, the hardware can now read whatever form of letter is presented in front of it. Using the neural network the success rate was found to be higher than 97%.