Abstract
Conference Title: 2019 International Conference on Wireless Technologies, Embedded and Intelligent Systems (WITS) Conference Start Date: 2019, April 3 Conference End Date: 2019, April 4 Conference Location: Fez, Morocco This work proposes a modified algorithm to give the Freeman code for the Arabic character (manuscript or printed, after making it thin) in a manner that takes into account all the forms on which the Arabic character can be, and with the lowest possible code length, without losing any information about the details of the character. So as to facilitate its later normalization and exploit it for the purpose of Arabic character recognition or for another purpose. We tested the code that our modified algorithm gives with a group of Arabic characters in its various forms and compared it with the code given by the algorithm used in previous related works. The average of reduction of our algorithm for the length of the code compared to the previous works algorithm is around 45%. In addition, the code we obtained represents the Arabic character accurately in all its forms, even those not represented by the previous works algorithm accurately.