INTELLIGENT SIGN LANGUAGE RECOGNITION SYSTEM USING SMART GLOVE

Abstract

Human-computer interface technologies aid speech- and hearing-impaired individuals to communicate. Body language is a great means to translate non-verbal gestures into language in real-time. In this wearable smart glove architecture, hand movements are converted to voice messages by the use of flex sensor-based motion detection. The resistive flex sensors on each finger detect the bending of the fingers in gestures. An Arduino microcontroller identifies specified patterns of gestures based on calibrated thresholding and mapping logic on analog data. The processed data of the gesture is transmitted to a mobile interface through HC-05 Bluetooth. The app has a text-to-speech engine that converts gesture labels into voice, and allows verbal communication to be smooth. Priorities are given to low-cost integration of hardware, low-latency wireless transmission, and real-time response in assistive communication. The modular construction allows users to create vocabulary of gestures that best fits their needs, increasing the flexibility in various contexts of use. An efficient voice synthesis architecture and a valid gesture differentiation offers non-verbal users an efficient alternative communication system in their everyday interaction.