A New Approach to Segmentation of Persian Cursive Script based on Adjustment the Fragments

Optical Character Recognition (OCR) is a very old and of great interest in pattern recognition field. The recognition of cursive scripts like Persian and Arabic languages is a difficult task as their segmentation suffers from serious problems in different languages. Segmentation is a process of dividing cursive words into smaller parts in order to decrease complexity and increase accuracy of recognition process. In this paper, an improved segmentation method of the Persian script has been presented and to increase the quality of segmentation, some structural features of Persian language is used to adjust the fragments. This method is robust as well as flexible. It also increases the system's tolerances to font variations. The proposed method is able to segment existing Persian fonts up to 99. 2% accuracy.