期刊名称：Journal of Emerging Trends in Computing and Information Sciences
电子版ISSN：2079-8407
出版年度：2015
卷号：6
期号：6
页码：268-274
出版社：ARPN Publishers
摘要：The increasing demand on digitization of human activities accompanied with developments in interactive technologies between human and computer, have led to an increase in interest of research for those who are focusing in the field of handwritten recognition of characters, words, sentences, and whole documents. The recognition task involves complex processes in artificial intelligence, image and signal processing. Semitic languages are different from European languages in many aspects including complex linguistic structure, implicit characters and concatenation, writing styles, fonts, and writing direction. The Arabic language, as one of the Semitic languages, has many unique characteristics that make the job of recognition even more challenging. Intensive research has been carried out in the recognition of handwritten English; however less effort has been paid for the recognition of handwritten Arabic. To this end, we propose in this paper applying a multi-level neural network for the holistic recognition of Arabic handwritten documents. In the presented methodology, many morphological operations are being applied followed by special objects recognition to localize and process handwritten words before matching them. The goal is to achieve high recognition ratios in short time. Therefore, the proposed recognition approach will be compared with the state of the art techniques in terms of accuracy and recognition speed.
关键词：Handwritten; Recognition; Neural Networks; Holistic word
