Abstract—In In recent years, fuzzy logic has been increasingly used to improve conventional methods especially in pattern recognition fields. This research present proposed system to recognize isolated Arabic characters, using a approach of fuzzy logic. The proposed system does not requires segmentation of the Arabic words or Arabic texts to characters but entered three different shape for every isolated Arabic character saved in many templates, after then applied laws and stages of fuzzy logic, save the results in different files, enter four shape for the character we wanted to recognize different from the three shapes were saved then compare the results recognition with the saved results of characters ,we used for testing the characters (Alif, Baa, Geam, Dal, Raa, Seen, Dhad, T'ta, Ain, Faa, Kaf, Lam, Mem, Haa, Waw, Yaa). Recognition results shown at many figures in the paragraph the experimental results for two different shapes Arabic isolated characters(Ain, Faa). The proposed system was tested on 96 different shape of characters and a 88% recognition success rate was obtained. The proposed system has been implemented and tested on Matlab R2008b environment.
Index Terms—Pattern Recognition, Fuzzy Logic (FL), Isolated Arabic Recognition (IAC), Template Comparison.
[PDF]
Cite: Sushil Garg, K. S. Kahlon, and P. K. Bansal, "Testability Analysis of Aspect Oriented Software,"
International Journal of Computer Theory and Engineering vol. 2, no. 1, pp. 119-124, 2010.
