Telugu Optical Character Recognition Using Deep Learning

Abstract

In a current generation many of researches developed to convert photogenic data into digital information. But this development consider only for popular languages like English, Latin and Chinese etc., but only few researches on religion languages. Coming to this project about conversion of Telugu handwritten data into digital form. It is not only printed conversion also handwritten data of different writing modules with different styles of writing. To convert into digital form here chosen two programming languages those are Matlab and Python. Considering the results and drawbacks finalise the python. It is mainly aimed towards to decrease time of user and increase the productivity and perfection outcome of data. An important features of this project is convert the Telugu handwritten into digital form using of pre-trained models also it works on offline.