Arabic Recognition Engine
The Arabic recognition engine only supports Arabic text with embedded numbers as well as French and English characters. No additional languages are supported. This recognition engine can use preprocessing that optimizes the image for character recognition.
As for any other recognition engine the results of the Arabic page recognition can be viewed in the Document Viewer. The Document Viewer is restricted to display recognition results from left-to-right for non-Arabic text, or right-to-left for Arabic text, it cannot display both. A text that contains at least 20% Arabic characters is considered Arabic, and the text direction is set to right-to-left even for embedded English or French text. This means that the recognized English or French words are then displayed in the reverse order and not as positioned on the document.