QuranVision

Abstract

Tajweed is a set of rules, which is required for every Muslim to learn in order to recite the holy Quran. These rules are needed to guide Quran’s reciters from making any errors, such as mispronouncing words which are strictly prohibited when reciting Quran. The conventional learning process on Tajweed rules, which comes in the form of face-to-face learning between instructor and the students, although considered the most effective way to learn Quran, may be tedious and time consuming due to the need for prolonged session of direct contact for learning. On that note, a mobile application which can detect and recognize the feature of Tajweed rules via image processing in al-Quran using a deep learning technology is proposed. Convolutional Neural Network or CNN is the commonly used technology when it comes to object detection and image classification. Model based on CNN architectures will be utilized in developing a real-time detection and recognition mobile application, focusing on Meem Sakinah-based Tajweed rules which include Idgham, Ikhfa’ and Izhar Shafawee. The key benefits of this application is in its ability to detect Tajweed rules in a real-time scenario and works in both black and white and color-coded Quran. This application provides an alternative way for new Quran’s learners from varying backgrounds and age to learn about Tajweed rules on their own time through visual and audio learning as well as detailed descriptions of the Tajweed rules which can aid in understanding the material more effectively.