الفهرس | Only 14 pages are availabe for public view |
Abstract The general aim of this research is to find new efficient solutions to unsolved problems that face Arabic handwriting segmentation and recognition such as Arabic ligatures segmentation, touching characters, overlapping characters, over segmentation, under segmentation, and misplaced segmentation. This research introduces an enhanced technique for Arabic handwriting recognition. The technique solves some unsolved problems that face Arabic handwriting segmentation and recognition. It presents a new morphological algorithm for the separation of Arabic ligatures into isolated characters. This research discusses a brief background of the system of Arabic handwriting segmentation and recognition. It presents an overview of most of the related work in this aspect. The methods used in each step and the problems that face each step are discussed. The stages of the proposed system are outlined in the following steps, Preprocessing, Parts of Arabic word segmentation, Liagtures segmentation, feature extraction, and Arabic character recognition. Another advanced Arabic handwriting recognition technique using lexicon reduction is introduced. The lexicon reduction technique stands on extracting the Arabic character shape descriptors. The technique implementation consists of the following major stages : extraction of the shape descriptor of each character. Arabic word Searching using Aho-Corasik string searching algorithm for Arabic character recognition. Various stages have been evaluated on the IFN/ENIT database. The results demonstrate the efficiency of the suggested techniques. The general aim of this research is to find new, efficient solutions to unsolved problems that face Arabic handwriting segmentation and recognition. The technique solves some unsolved problems that face Arabic handwriting segmentation and recognition. |