Classification Arabic and Dialect Iraqi Spoken to Sign- Language Based on Iraqi Dictionary

Sign language(SL) for the deaf or hard hearing is the best way to communicate between their community and other communities. Many researchers have dealt with this topic due to its utmost importance for dealing in creating an easy and accurate way to communicate with deaf groups. There are many computer technologies that have been used to deal with SL and have given the most importance to how find a high-tech and accurate way to convert text or audio into SL. In this study, the Arabic sign language (ArSL) was dealt with, especially the Iraqi sign dialect because it lacks a system or application to deal with SL like the other Arab countries, many studies deal with this problem and several techniques depending on the amount of data deal with this groups. The aim of this study deal with Iraqi -data Dictionary used in deaf schools. In this paper used the decision tree (DT) by machine learning (ML) algorithm where DT is the best technique use with voice recognition specially in Arabic language to classify the spoken Arabic word to SL as image, also the objective to reach this classification by using multi techniques (noise reduction, filtering,…) to achieve a high accurate deal with deaf human in easy communicate. This study applied on 25 person's cases, then analyzing their voice through use a set of features that help to convert audio into text and then into ArSL, indicating the language of Iraqi dialect, and adopting the data of the Iraqi signal dictionary, and Compared to two researches used Convolution Neural Network (CNN), Recurrent Neural Network RNN methods, found this proposed system result achieved the best an accuracy rate of about 94%.