A hand gesture recognition based communication system for silent speakers

In this paper, a communication system has been proposed which converts sign language used by dumb people into speech. It is done based on the novel hand gesture recognition technique. This solution approach consists of a hardware module and software application. In hardware module - The Gesture recognition is done with the help of sensor glove which consists of 5 accelerometer sensors, a microcontroller and Bluetooth chip which are best positioned in fingers, based on the analysis of American Sign Language (ASL) signs. The design of glove and the concept of decoding gestures by considering the axis orientation with respect to gravity and their corresponding voltage levels are discussed. In Software part - an android application named Speaking gestures have been developed. It receives the data (alphabet/word) via Bluetooth, converts them into text and speaks it out. The entire process of speech synthesis has been tested and the test results displaying the alphabets and words have been shown.