Text Identification System for Translation of English Language

Abstract

India is a multilinguistic country. People of different states speak different languages but all Indians are not polyglots. English is called as universal language and Kannada is one of Dravidian languages that is used in India. The majority of people in India, especially who live in villages, cannot read, write and understand English language. Translation of any language that is done by human is time consuming. Therefore, there is a need for implementation of Machine translation systems that will translate English language to Kannada language. This paper describes translation system that enhances the knowledge of society without language barrier by translating text in English language to Kannada language by giving the meaning of the word. We can click the picture of any English text and give Image as input to translation system. The system translates English text to Kannada text by extracting characters from an image by using tesseract software. Tesseract is optical engine that scan image containing text and extract characters from image and convert them into editable from. Extracted text which is in English language might have non dictionary words that may or may not be translated properly to Kannada language by existing system. It also identifies such non dictionary words compares with hybrid database and replaces these words with exact Kannada word.