Using group of statistical operator toward Optimal segmentation of Arabic word signal to its basic phonemes

Abstract

In this research, we are segment the sound Arabic signal to its basic group of phonemes depending on three statistical operators that are( mean ,standard deviation ,covariance) to reduce the error in determining the correct points segment . Each operator is implemented on the data of the sound entered word and then the results are manipulated by some processes that give intermediate results , these results (from each implementing of operator) merges together and give union results that are passed to postprocessing in order to prepare to find the points segment in high efficient.