Global Journal of Information Technology: Emerging Technologies

Yazarlar: Mohamad Reza Keyvanpour, Reza Azmi, Zahra Sadat Mortazavi Tabatabai, Zahra Abdolhosseini

Konular:

Anahtar Kelimeler:Handwriting Persian Character Recognition,Genetic Algorithm,Orphological Operation,Tructural Elements.

Özet: Handwriting recognition refers to the identification of written characters. It is the task of determining what characters or words are presented in a digital image of handwriting text. Also, it is sub-task of Optical Character Recognition (OCR).  Recognition of handwritten Persian characters is complicated because of the similarity between characters and the different styles of writing and this process requires bunch of algorithms and methods. In this paper we introduce a novel approach to recognize handwriting Persian characters. We used morphological operators to describe each character structure and applied genetic algorithm to structural elements optimization then these are used for classification unseen data. We test our algorithm using a dataset of hand-printed Farsi characters which includes samples for each 33 hand-printed characters. Experimental results show that proposed method improves the recognition performance.

---