Printed Arabic text recognition

الملخص EN

-This paper presents two printed Arabic text recognition systems.

Both of them use the same text segmentation procedure which segments the text effectively with a rate of 98 %.

The first system employs a syntactic-structural recognition method.

It has a large tolerance suitable for multifont recognition, and a recognition rate of about 90 %.

The second system is based on the principle of normalization and similarity detection.

It has a 99% recognition rate and the ability to learn new Arabic typewritten fonts rapidly