Automatic Sign Language Gesture Recognition using Prewitt & Morphological Dilation

Abstract

, Sign languages have their own linguistic structure, grammar and characteristics, and are independent of the rules that govern spoken languages. They are visual languages that rely on hand gestures as well as on bodily and facial expressions. Sign languages in different countries are vastly different from one another, so enabling easy communication is important: not just to break the barrier between hearing and deaf individuals, but also between people who do not sign in the same language. In India, sign language plays an important role in the field of communication among dumb and deaf people. There are different signs associated for communication in every country as per their convenient gestures. Automatic sign language gesture recognition is an approach for recognizing gestures and converts it to its actual meaning and convey either through speech or text as per requirements. Here the system is based on Prewitt Edge Detection that possesses the gestures of sign language and helps to recognize and assign their meanings. The Prewitt is second order derivative that has been used in image processing and computer vision, in the form of edge detection or extraction algorithms where it creates gradient of horizontal and vertical magnitude. System also uses certain pre-processing filtration technique such as morphological dilation for better feature extraction.