Sign language interpreter hand using optical-flow

Abstract

Communication is needed to interact and to socialize. In general, communication is used verbally, that is to use spoken or written words. But in everyday life, some of people are unable to communicate verbally. Some of them have a problem in communication, such as deaf and mute. The most effective communication used by deaf and mute are through non-verbal communication. This communication uses sign language in the form of hand and body gestures. Hand gestures identification is one way to identify the position pattern of the hand and the body in order to translate the meaning and the purpose. This communication is intended so that a normal person can understand the sign language which is conveyed by the deaf. In understanding the sign language, a communication media is needed to interpret sign language into easily-understood verbal language. The sign language described in this research only includes sign language from hand and body gestures which then will be translated into characters. The important stages and processes that must be done in the implementation of sign language system is to detect the configuration of hand gestures which describe the Indonesian Sign Language (BISINDO). The method used in this research is the Optical Flow that is used to detect hand and body movements as the object with the red glove as a marker. Based on testing conducted on 20 different types of word recognition results obtained : 65,5% identified for 20 species of the word.