Android based Sinhala Document Reader for Visually Impaired Persons

Abstract

Visual impairment is a special kind of disability found around the world. Governments in different countries try many approaches to make visually impaired people s life easier. Going with that convention, Sri Lankan government also try to identify them as differently abled people aiming to fill the gap between regular people and them. There are two major types of people found with visual disability; partially disabled, and fully disabled. Reading is very important activity in learning, but it is an example which both types of these people found difficult to perform. From technical aspects, there are many solutions implemented for this reading activity problem. Computer aided braille systems, text-to-speech systems are two examples for such applications. Being a third world country, everybody in Sri Lanka cannot afford a computer based solution due to its high price tag. Unavailability of Sinhala language based systems is another reason for computer based solutions to become not popular within country. Mobile devices are very powerful toys nowadays due to their high capacity and processing power. Also their price has been reduce significantly in the last couple of years due to market competition, especially among Google Android based devices. For real time Optical Character Recognition and Text-to-Speech systems, such devices shows a great potential due to their portability. Combining such systems into a mobile device will give an acceptable solution to visually impaired persons because they can capture the documents, and then listen to the text of the document by couple of finger touches. With the introduction of accessibility features in Android devices, a blind person should be able to learn to execute an application using the touch interface of the device, which makes this solution more practical. Sri Lanka can also take the advantage of such mobile systems due to the high popularity of mobile phones within the country, which has been estimated to be equal to the current population. Also there are people who like to donate such a device to a blind person, if there are applications available .This research paper addresses such a solution of Sinhala language based Optical Character Recognition system and a Document Detection system by incorporating it into Google Android powered devices. Also the system will assists the user by audio notifications to correctly frame on a document and facilitates to automatically snap it and process to Unicode based characters. By incorporating with a Sinhala based Text-to-speech system, the user should be able to listen to the text that he has captured in Sinhala language. This solution will be a major life changer of visually impaired people.