Image to text conversion in foreign language using document image processing technique

Abstract

Every country has their own native languages such France has French, Japan has Japanese, India has Hindi and other local languages like Gujrati, Marathi, Telugu, Tamil, Bengali, etc. When a person who does not know the English Language travel to a country whose local language is English will face many problems like understanding road sign texts, shop’s name, Instructions Boards etc. So, in this paper we are proposing a model with the help of which user can take snap of the scene containing text of which he/she want to translate it, upload that photo, choose the language and the software will provide us the output in text of User’s native language. This model will use Convolutional Neural Network (CNN) to identify the characters, Efficient and Accurate Scene Text Detector (EAST) for detecting the text in the image, Digital Image Processing methods are used to segment the text in the detected text, googletrans for translating the text to other language and Tkinter for creating Graphical user Interface (GUI) for software.