Low-cost intelligent gesture recognition engine for audio-vocally impaired individuals

This paper presents a low-cost open source prototype of a gesture recognition and interpretation glove that aids audio-vocally impaired individuals in communication. A sensor based glove for recognition of the gestures in the American Sign Language (ASL) is the input module to the system. It incorporates resistive flex sensors and contact sensors appropriately positioned to obtain a theoretical efficiency of 96% in character recognition. When the sensors are subjected to change in orientation by the user, the variation in electrical resistance of each of the individual sensors is obtained as the input to an Arduino ATMega328 microcontroller that accurately maps it to the letters of the English alphabet through an efficient algorithm. An elegant methodology that is easy to implement using a microcontroller is employed for the mapping process to minimize the complexity of the system. The paper also presents a detailed study of the efficiency characteristics involved in the construction of the custom-designed flex sensors utilized in the system. The paper outlines the calibration and testing results of the sensors .The highlights of the system are its ergonomic and economical design, portability and intelligent gesture recognition and interpretation ability. The implemented system achieves a practical efficiency of up to 80% of theoretical efficiency. The cost of the system is approximately less than USD 5 at laboratory condition using off the shelf components.