Tajweed checking system to support recitation

Abstract

Al-Quran recitation has a set of rules known as tajweed to ensure proper pronunciation, readings and interpretations of the Al-Quran. It has been traditionally taught by ustad or experienced religious teachers. Generally, these teachers will listen to the students recitation and highlight their mistakes. However, the traditional method which requires the present of these skilled teachers has limitation to support self-learning environment. For this reason, we propose tajweed rule checking system using speech recognition technology to help students to learn and revise proper Al-Quran recitation by themself. The proposed system is capable to recognize and point out the mismatch between the students recitation with the recitation made by the experienced teachers stored in a database. The system adopts the MFCC algorithm for feature extraction and HMM for feature classification.