PEMBACAAN GERAK BIBIR UNTUK PENGENALAN PENGUCAPAN HURUF HIJAIAH MENGGUNAKAN HIDDEN MARKOV MODEL (HMM); LIPS MOTION READING FOR RECOGNITION HIJAIAH PRONUNCIATION USING HIDDEN MARKOV MODEL

SANTOSA, HARDI, Triyogatama Wahyu Widodo

2016 | Skripsi | FMIPA

Recognition the pronunciation of hijaiah is important, because it can be used as a basic for studying the Quran. Indonesian is a country that is predominantly Muslim, but the language used in everyday use alphabet. To facilitate the study of the Quran which uses the Arabic alphabet, it needed an auxiliary media. In this study built a speech recognition system of hijaiah pronunciation based on lips motion. Letters hijaiah identified only certain letters, the letters that make up the letters muqoththoah. This System used haarcascade Features of the Mouth. The feature is used to obtain the coordinat key point of the lip. The resulting coordinat key points of the Lip extracted to determine The characteristics of The Lip are value centroid (lips feature). Value of Centroid is used in the recognition process. Recognition process is done using Hidden Markov Model. The results showed an introduction The pronunciation of hijaiah based on The pattern lip has good accuracy in the testing of the letter “Lam”, “Ro” and “Sin” with an accuracy of 84 % and The lowest of introduction on The letter “Ain” with an accuracy of 80%.

Kata Kunci : Hidden markov model, Snake (active contour), hijaiah, muqoththoah.