Recognition of Farsi Letter using Hidden Markov Model

Sadegh Zarezade, Abbas Akkasi, Ayoob Maher, Yalda Namdar
2015 International Journal of Computer Applications  
Letter recognition is taken from optical character recognition (OCR). Some of the applications like devices which read postal code and checks are limited to recognition of numbers and need high speed and accuracy. In current paper the combination of two powerful method i.e. hidden Markov model will be used. Other models are used only in recognition of English words in learning using online method. The accuracy of recognition is 93% for Ifn/Farsi Database.
doi:10.5120/ijca2015905791 fatcat:ufmgt43bsvexdo5wckkunyu2lm