Journal of Computer Science

Multilevel Classifier in Recognition of Handwritten Arabic Characters

Rawan Ismail Zaghloul, Enas Faisal AlRawashdeh and Dojanah Mohammad Kadri Bader

DOI : 10.3844/jcssp.2011.512.518

Journal of Computer Science

Volume 7, Issue 4

Pages 512-518


Problem statement: Arabic offline handwriting recognition is considered one of the most challenging topics. This is probably caused by the fact that Arabic recognition system faced many problems during the development stage. It faces the usual problems of character recognition in general, in addition to the problems that are specific to Arabic language only. The aim of this study was to build a classifier to solve Arabic text ambiguity; to be used in text recognition applications. Approach: A multilevel classifier based on pattern recognition techniques, is proposed. The proposed classifier was implemented using MATLAB and also tested with a large sample of handwritten datasets. Results: Pattern recognition techniques are used to identify Arabic handwritten text. After testing, the recognition rates reached {93, 84, 89 and 85%} for the isolated letters, letters at the beginning, at the middle and at the end of the word respectively. Conclusion: Even that the Arabic letters change their shape depending on their position in a word, the proposed classifier, using the powerful set of features, is proved to be effective in the recognition of Arabic letters.


© 2011 Rawan Ismail Zaghloul, Enas Faisal AlRawashdeh and Dojanah Mohammad Kadri Bader. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.