• Ahmed Hamid Saleh Northern Technical University / Administrative Technology Collage, Mosul


The letters form a front face that can be recognition as a transformative process that transforms the paper document into a fully editable form. The explanatory necessity because of the urgent need associated with the work of the computer, which is based on the processing of the data, shortens the time and limits it based on the accuracy and speed of completion of the computer work as well as clarity by the user.
This paper suggests a new method of letter recognition based on modified the technique of threshold and correlation coefficient, which marks the process of recognition between the 28 letters and letters of the Arabic language. The results obtained in the applied process on the set of these characters in various and multiple sizes(volume letter) (24, 26, 28, 30, 34, 38, 40, 45, 50) and multiple lines (Simplified Arabic, Arial, Traditional Arabic, Tahoma), which effect obtaining high scores when calculating the recognetion ratio of 99.70, this method has the flexibility and speed of other methods used to process this process.


Cheriet, M., N. Kharma., (2007). Character Recognition Systems: A Guide for Students and

Practitioners, Wiley.

Srihari, S.N., A. Shekhawat, and S.W. Lam. (2003), Optical character recognition (OCR).

Ahmed, T., Mohammed, R. and Abdelhay, S. (2014). Investigating of preprocessing techniques

andnovel features in recognition of handwritten Arabic characters. Lect. Notes Artif Int., 2014; 264-276.

Rawia I. O. Ahmed و Mohamed E. M. Musa. (2016). Preprocessing Phase for Offline Arabic

Handwritten Character Recognition.

Dehghan, M., Faezl, K., Ahmadi, M. and Shridhar, M. (2000). Off-line unconstrained

Farsihandwritten word recognition using fuzzy vector quantization and hidden Markov word models .In:

IEEE 2000 Proceedings of the 15th International Conference on Pattern Recognition 3-7September 2000;

Barcelona, Spain: IEEE. pp. 351–354.

Amir, M, Karim, F. and Abolfazl, T. ( 2002). Feature extraction with wavelet transform

forrecognition of isolated handwritten Farsi/Arabic characters and numerals. In: IEEE 2002International

Conference on Digital Signal Processing; 1-3 July 2002; Santorini, Greece,IEEE. pp. 923 – 926 .

Amrouch, M., Elyassa, M., Rachidi, A. and Mammass, D. (2008). Off-Line Arabic

handwrittencharacters recognition based on a hidden Markov models. In: IEEE 2008 International

Conferenceon Image and Signal Processing; 1-3 Jul; Cherbourg, France: IEEE. pp. 447 – 454 .

Ahlam, M., Akram, H., Khalid, S. and Hamid T.,(2014). Using HMM toolkit (HTK) for recognitionof

Arabic manuscripts characters. In IEEE 2014 International Conference on MultimediaComputing and

Systems; 14-16 April 2014; Marrakech, Morocco: IEEE. pp. 475 – 479.

Mahmoud SA. 2008. Arabic (Indian) handwritten digits’ recognition using Gabor-based features. In:

Innovations in Information Technology, 2008. IIT 2008.International conference. Piscataway: IEEE, 683–

Abdul-Azim, El-Sherif E. 2008. Arabic handwritten digit recognition. International Journal of

Document Analysis and Recognition,11(3):127–141.

Safaa S.O. Jumana A.J. Iraqi car license plate recognition using ocr. annual conference on new trends

in information and communication technology application. Iraq.2017.

Otsu, N., (1975), A threshold selection method from gray-level histograms, Automatic.