DETEKSI KETEPATAN PEMBACAAN SURAH AL-KAUTSAR MENGGUNAKAN EKSTRAKSI CIRI MEL-FREQUENCY CEPSTRAL COEFFICIENTS (MFCC)
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Telkom Purwokerto
Author
ADITYA, RACHMAN
Subject
T Technology (General)
Datestamp
2021-04-23 06:47:53
Abstract :
ABSTRACT The accuracy of reading the words in each verse of the Qur'an is the most important part of the Muslim murottal holy book. Errors can occur in several aspects such as the lengths of reading or pronunciation of words. In this thesis, an auxiliary system has been made to detect the accuracy of Al-Kautsar recitation with the test parameters in the time and frequency region. This final project performs word extraction and tests the accuracy of reading AlKautsar surah. The feature extraction method used is Mel-Frequency Cepstrum Coefficients while the detector (classifier) is the Cosine Similarity method. The results of system testing with a threshold of 0.57 to 0.6 obtained the percentage of 100% word recall with 100% word precision for correct reading training data, for incorrect reading obtained 100% word recall with 100% word precision. In the correct reading test data obtained 100% word recall with precision words 97.83%, while for the wrong reading obtained 100% word recall with the word 99.26% precision. MFCC feature extraction method is good enough to detect the accuracy of reading surah Al-Kautsar with correct recitation. Keywords : Voice, Voice Recognition, Al-Kautsar, MFCC, Cosine Similarity .
ABSTRACT The accuracy of reading the words in each verse of the Qur'an is the most important part of the Muslim murottal holy book. Errors can occur in several aspects such as the lengths of reading or pronunciation of words. In this thesis, an auxiliary system has been made to detect the accuracy of Al-Kautsar recitation with the test parameters in the time and frequency region. This final project performs word extraction and tests the accuracy of reading AlKautsar surah. The feature extraction method used is Mel-Frequency Cepstrum Coefficients while the detector (classifier) is the Cosine Similarity method. The results of system testing with a threshold of 0.57 to 0.6 obtained the percentage of 100% word recall with 100% word precision for correct reading training data, for incorrect reading obtained 100% word recall with 100% word precision. In the correct reading test data obtained 100% word recall with precision words 97.83%, while for the wrong reading obtained 100% word recall with the word 99.26% precision. MFCC feature extraction method is good enough to detect the accuracy of reading surah Al-Kautsar with correct recitation. Keywords : Voice, Voice Recognition, Al-Kautsar, MFCC, Cosine Similarity .
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
