INTERPRETATION OF INDIAN SIGN LANGUAGE USING STORED VOICE SAMPLED VALUES

Abstract

According to the census of India 2011 data on disability, around 7.4% of population is speech impaired.
They use sign language as a medium to communicate but it restricts them to communicate with people who do not
understand sign language. Speech impediment not only restricts them in communication but also in learning. This
work make its contribution towards bringing down the barrier of communication and thus developing a compact
and affordable wearable device which converts Indian sign language to voice and text. It is equipped with a
accelerometer weared in index finger. For each gesture, accelerometer yields XYZ coordinates which are sensed by
ATmega 2560. According to the sensed coordinates micro-controller extracts the corresponding speech file
(decimal sampled sound file) from SD Card, process it and feed to audio amplifier and at the same time to LCD
display. The .wav files are processed via Matlab to fetch the decimal sampling of sound files which are stored in the
SD card for memory management and hence extends the potential for vast storage of words and sentences