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 sound which share approximate significant chosen for study ability distinguishing among them through abstracting characteristic them. The signals recorded microphone represented in a binary matrix. This procedure was implemented prepare these processing operation two co-occurrence matrix (contrast, energy) counted. values were compared from one person another discover properties sharing common their articulation another. results analysis gave dependence distinguish speaker, addition high provided arabic letters, where no connect between both elements signaling any letters.