International Science and Technology Journal

Published by

Under supervision of


Open Access Journal

ISSN: 2519-9854 (Online)

ISSN: 2519-9846 (Print)

DOI: www.doi.org/10.62341/ISTJ

A peer-reviewed and open access journal concerned with publishing researches and studies in the field of applied sciences and engineering

Published by

Under supervision of

Intelligent Tifinagh Alphabet (ⵜⴰⵎⴰⵣⵉⵖ) Recognition System Using Back Propagation neural network BPNN

Researcher(s):
  • Ahmed Said Ahmed Ashit
  • Salam Adeam Abdulah Fatnasa
  • Nabil Abulgasem Ali Abrebish
  • Souad Ali Omer Burak
Institution:High Institute of Science and Technology Alharaba, Libya
High Institute of Science and Technology Alharaba, Libya
IT department, university UUM, Malaysia
High Institute of Science and Technology Alharaba, Libya
Field:علوم الحاسوب و النظم الخبيرة وتقنية المعلومات
Published in:13th volume - March 2018
الملخص
الحروف الأبجدية التيفانية استخدمت من قبل الامازيغ في شمال إفريقيا وجزر الكناري حتى القرن الثالث عشر . في هذه الورقة تم استخدام الشبكة العصبية ذات التغذية العكسية للتعرف على الحروف الأبجدية التيفانية , حيث تم تحجيم صورة الحرف ليكون ملائم للشبكة Back Propagation neural network . في النظام المقترح تم تمثيل صورة الحرف بالأحرف الثنائية (0,1) . كانت نسبة التعرف للشبكة Back Propagation neural network حوالي 79% .
Abstract
The Tifinagh alphabet in order ("Lybico-berber") has been utilized by Amazigh talking individuals in North Africa and the Canary Islands in any event from the third century B.C. up to the third century A. In the present paper, artificial neural network is used to the to recognize the Tifinagh alphabet which used for writing Amazigh language. In this paper the alphabet image is reshaped to be suitable as input for networks, wherefore, image processing resizing technique is used. The proposed method is based on the use of feed forward back propagation method In the proposed system, alphabet's image is represented by binary numbers that are used as input, then they are fed to the neural networks. Neural network followed by the Back Propagation Algorithm which compromises Training. The recognition rate for Back Propagation neural network BPNN was 79% in testing phase.