تمييز الأرقام العربية المفردة النطق باستخدام الخوارزمية الجينية

Abstract

This study aims at constructing an intelligent system for recognizing the single Arabic numbers. It consists of two basic stages: the stage of features extraction and the stage of recognition. In the first stage, the technology of (Mel-Frequency cestrum coefficient (MFCC)) was employed. But in the second stage, the genetic algorithm was used. The results of the test showed that words recognition percentage was (100%) for the words used with training, and it was (97%) for the words used with no training. The proposed system was constructed using the Matlab version (0.7) program, and the data used in the system are the following numbers: (0, 1, 2, 3, 4, 5, 6, 7, 8 and 9). Also, six speakers (four males and two females) performed the voice recording.