Enhancing Fuzzy Logic for Pattern Recognition Technique to Recognize Arabic Quran Short Diacritics

Dr. Majdi Salameh; Dr. Suleiman Almasri

A word in Arabic language with different diacritics gives different meaning and pronunciation. It is important to recognize Arabic characters and diacritics for accurate reading; writing and pronunciation. Significant research interest about Arabic text recognition problem was attracted in the past. Recognition problems usually emerged as diacritics and ligature. This paper uses effective recognition approaches that enhanced the recognition of diacritics by developing several algorithms and techniques to classify and recognize Arabic and Quran diacritics. Proposed technique recognizes double, triple, and special diacritics using fuzzy logic for pattern recognition. The technique recognizes all Arabic diacritics including diacritics in many Arabic font types and diacritics in all positions of Arabic text. This paper deal with diacritics as a separate part of the text. Arabic diacritics possibly could be complex, as in the special diacritics for the holy Qur'an. Arabic diacritics are recognized using fuzzy logical for pattern recognition technique which is basically uses strokes which in its turn compost diacritics, in order to calculate the lines angle using polygon formula to be determined as one of the main lines and curves. Afterwards, stores strokes in vectors to represent all diacritics. Then the unknown diacritics vector is compared with all stored vectors to determine the correct diacritics. Recognition results for special and normal diacritics using fuzzy logic for pattern recognition adopted technique scored 97%, and for normal diacritics scored 94% because of the missing of classifying. Over all diacritics recognition scored the result scoring 95.6 %.

Select Volume / Issues:

• Volume 8,Issue 1, Jan. 2017

Year:

2017

Type of Publication:

Article

Keywords:

Arabic Text Recognition; Diacritics Recognition; Quran Recognition; Fuzzy Logic; Pattern Recognition; Short Diacritics

Journal:

IJECCE

Volume:

8

Number:

1

Pages:

62-68

Month:

January

ISSN:

2249-071X