Recognition of Off-line Arabic Handwriting Using Hidden Markov Model Toolkit

This paper presents an off-line Arabic handwriting recognition system using the Hidden Markov Model Toolkit (HTK). HTK is a portable toolkit for speech recognition system. The recognition system extracts a set of features on binary handwritten images using sliding widow, builds character HMM models and learns word HMM models using embedded training without character presegmentation. Moreover this paper studies the relationship between frame overlap and number of stats. Experiments that have been implemented on the benchmark IFN/ENIT database show the average recognition rate of this system is 85.43%.