Tajweed Segmentation Using Pattern Recognition, Extraction and SURF descriptor Algorithms

Abstract

This paper proposes a method of detection and recognition algorithms and recitation of the Qur'an. The methods and algorithms are packaged into a software system application to allow a person to learn recitation and how pronunciation is good and right and real-time. Any recitation of the Qur'an written in a different color of the letters and punctuation instead of recitation and tajweed each have different shapes. Variable color is used as a reference for the segmentation process forms of recitation and SURF algorithm used for feature extraction process forms. Feature any form of recitation is stored in the data base (knowledge base) accompanied by explanatory text recitation of data and audio pronunciation files. In the process of recognition when the user enters a query tajweed image, Euclidian distance is used to measure the similarity between the query and the shape feature tajweed recitation shape features that exist in the database.