Character recognition systems have contributed tremendously to the automation of data entry processes. Arabic characters are widely used in Arabic countries. If optical character recognition (OCR) systems are available for Arabic characters, they will be very useful and will have a high commercial value. Therefore, we have implemented a statistically based Arabic OCR system and present the techniques used in this paper. It consists of image acquisition, preprocessing, segmentation, feature extraction and statistical classification. The developed system is window-based software that can interact with a scanner and can recognize a single font of Arabic script with 85% accuracy at about 16 characters/second.
展开▼