Design and development of hand gesture recognition system for speech impaired people

Abstract

Communication is a vital tool necessary among all the people of the world and available in many forms such as body language, telepathy, sign language and Morse code. Sign language is an effective method of communication for speech-impaired people, who require an interpreter for the purpose of conveying information. The sign language recognition is the important research field for the period of last 10 decades. Sign language is not an universal language all over the world. Each country has its own sign language such as American sign language (ASL), British sign language (BSL), Indian sign language (ISL), Chinese sign language (CSL) and etc. newlineISL is an exclusive sign language consisting of head movement, face expressions and both hands to perform a gesture. ISL is a medium of education used in the speech impaired schools all over India. In this study, the gesture recognition system is designed and developed uniquely for Indian sign language (ISL) recognition. The two most dominant approaches for sign language recognition were sensor-based and image-based. The ISL gestures database have been developed in both approaches and classified. newlineIn the sensor-based approach, sensors have been fixed on the double handed gloves. This glove is termed as Sign and Sound glove (SS glove), which have been designed for three different datasets of single handed words, characters and double handed words. The sensors employed were flex sensor and accelerometer on the fingers and wrist position respectively after calibration. The flex sensor of 2.2 inch length would capture the bending of gestures. The rotation, orientation and inclination of the hand shapes were obtained using 3-axis accelerometers. newline