A HTK-based System to Recognise Arabic Script

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


Document Analysis, Pattern Analysis and Recognition, Machine Vision, Arabic OCR, HTK


This paper presents a cursive Arabic script recognition sys tem. The system decomposes the document image into text line images and divides each text line image into smaller overlapped frames. The system extracts a set of simple sta tistical features from each frame and then injects the se quence of the feature vectors to the Hidden Markov Model Toolkit (HTK). HTK is a portable toolkit for speech recog nition system. The proposed system is applied to a sophis ticated cursive Arabic font.

Important Links:

Go Back