Abstract
The sign language is the primary means of communication among the deaf community and normal people. It is the natural language for the community, using hand movements and other gestures. Many attempts are being made to develop an Arabic Sign Language (ARSL) and distribute it among members of the deaf community and those concerned. Building an accurate sign language recognition system is of a great importance in order to efficiently facilitating this communication. In this research, an Avatar Based Translation System for ArSL (ABTS-ArSL) is developed using Microsoft Window 7 Speech Recognition Engine. A proper training on the Speech Recognition tutorial is suggested prior to using ABTS-ArSL to achieve better voice recognition. ABTS-ArSL grammar is developed and embedded in Microsoft Visual Basic 2008. The Arabic alphabets and single numbers with their respective signs and avatar images are stored in different directories and invoked upon receiving a word uttered by a user. ABTS-ArSL is a visual natural language model that might be used for communication by both groups in Arabic speaking countries.