Arabic vowels characterization and classification using the normalized energy relating to formants and pitch

Vowels are the primary units of a sound system language. The classification these vowels is therefore very important for recognition and synthesis speech. In this paper, we propose normalized energy-based approach in formants pitch to characterize Arabic (short vowels: / /, i u /; long a: i: u: /). was performed using developed algorithm on records extracted from an corpus after extraction first three computation energy bands. results showed that distinguishes by analyzing nucleus F1, F2, F3 F0 with rate 88.7% 90% short vowels.