Auto electronic recognition of the Arabic letters sound

In this research Arabic speech sounds have been studied and investigated, so as to find the distinctive features of each articulated sound. Therefore, certain Arabic sound which share certain approximate distinctive significant features have been chosen for study the ability of distinguishing among them through abstracting characteristic features for them. The signals of speech for the sounds have been recorded through the microphone which represented in a binary matrix. This procedure was implemented so as prepare these signals for processing operation through which two features for the co-occurrence matrix (contrast, energy) have been counted. The values of these features were studied and compared from one person to another to discover the certain speech sounds properties sharing certain common distinguishing features approximate in their articulation one another. The results analysis for this study gave the ability of the dependence to these features for distinguish the sound of speaker, in addition to the high ability which provided to distinguish among the arabic letters, where no connect between both co-occurrence matrix elements and the features of signaling of any arabic letters