Handwriting Segmentation of Arabic Text

R. Bentrcia and A. Elnagar (UAE)


Character segmentation, Arabic text, intelligent agents, character recognition.


The segmentation of words into characters is a main stage in character recognition systems. In this paper, a novel approach is proposed to segment Arabic words, written in Naskh handwriting style. The segmentation algorithm is based on seven agents which cooperate to detect regions where segmentation is illegal. Then, end point features are extracted from the remaining regions of the word and the middle of every two successive end points is considered as a candidate segmentation point if specific rules are satisfied. The experimental results are very promising and achieve a success rate of 86%

Important Links:

Go Back