Deep Learning Design For Character Recognition With Position-Free Touchscreen-Based Braille Input Method
S. SHENBAGAVADIVU · i-manager s Journal on Mobile Applications and Technologies · 2021
Specially challenged people means they would have different abilities or some sort of disabilities when compared with the average people. With the technology advancement, almost all the impairments are getting addressed by using smart devices. Braille is a method used by visually impaired people to communicate with others. Smartphone apps which get Braille input and convert it into natural language is the most beneficial invention to visually impaired people. But the touchscreen of the existing system is a location-dependent method in which blind people have to place the Braille dot at a specific location. It may be a difficult and tedious process to know the location. The proposed system focuses on position free touchscreen method which would be easy to place dots anywhere on the screen. This is an innovative Braille input method using smartphone. Datasets were trained and tested in three languages such as English, Tamil and Hindi. Then by using deep learning techniques the character is identified and the same is given as audio output. This paper is useful to visually impaired people who want to use smartphone with position free touchscreen method. Thus, the proposed system is a new Android smartphone system which is most effective and useful for visually impaired people.