Abstract
Optical Character Recognition (OCR) systems were developed with high accuracy to facilitate transactions and increase human-computer interaction in most levels of government and commerce sectors. OCR has been adopted for diverse languages, but a few efforts have mainly been conducted in Arabic characters, and it has suffered weakness in the Arabic language. Therefore, a new Arabic Optical Character Recognition (AOCR) system is proposed to achieve a highperformance recognition system in printed images. Several steps are conducted to achieve the proposed AOCR system, such as image preprocessing, segmentation (line, words, and character), feature extraction, and classification. After evaluated the recognition system with multi-criteria, the AOCR results have shown accuracy with 95% with different quality document images (spatial resolution); besides, the system was adequate to resist the degradation of the documents, compared to other commercial systems in the literature.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
