Optical Character Recognition of Devanagari Script Using Machine Learning-A Survey

Abstract

Optical character recognition(OCR) became the powerful tool when it comes to most of the digital world applications, there are wide variety of languages and script styles throughout the world when it comes to countries like America,Russia,Europe the script is almost identical or similar and the research has been started few decades ago and they developed efficient algorithms to any printed document or content on the image or the content on the handwritten document to editable text on a digital device. The beauty of India is it has one basic language that is Devanagari lipi (script), many languages like Hindi, Telugu,Malayalam,Tamil,Kannada from different parts of the county consisting of similar kind of characters which are from devanagari character set , due to this similarity of the character styles the existing algorithms which are efficient for foreign languages are identifying the wrong letter from the printed document or hand written document, to overcome this problem in this problems in this paper mainly we are concentrating to do survey of the work that has been done over few decades in Devanagari fonts identification using Neural Networks.