Handwritten english character recognition and speech synthesis to aid text-to-speech applications

Abstract

Handwritten character recognition is principal area for research. This domain is widely challenging. It has various applications like helping the visually impaired and in government offices where there is a need to recognize handwritten documents. Through this paper we propose a system which recognizes characters and converts them into voice output using text-to-speech system. Datasets of uppercase letters, lowercase letters and digits have been used as templates for matching with the input image. every input image of the character that we have considered is of the standard size 42x24 pixels. Our proposed system manages to produce good recognition rates and successful voice output.