An Omni-font HTK-based Arabic Recognition System

S.A. Al-Qahtani and M.S. Khorsheed (Saudi Arabia)


: Machine Learning, Man-Machine Interfaces, Computer Vi sion, Image Processing, Arabic OCR, HTK


This paper presents an omnifont, unlimited vocabulary recognition system. The system does not require segmen tation, and it is applied to cursive Arabic script where lig atures, overlaps and style variation pose challenges to the recognition system. The system is considered as the first at tempt to recognise Arabic script using the Hidden Markov Model Toolkit (HTK). HTK is a portable toolkit for speech recognition system. The performance of the proposed sys tem is assessed using a data corpus that includes six sophis ticated computergenerated fonts.

Important Links:

Go Back