Arabic number recognition

I am trying to detect Arabic numerals (Arabic - sign) from an image.

Tried Tesseract OCR, it didn't work for me (it recognizes Arabic words but not numbers) here is the image I would like to extract the page number from it (top of the page).

enter image description here

I tried imagemagick to compare this image with the smaller already made small images that contain all the book number possibilities, but that didn't work either and I think it will take so long.

What would be a practical non-complex solution? PS: The image will be from android phones and will be parsed on windows or linux server.

+3


source to share


1 answer


Actually Tesseract is not a valid solution to your problem or any commercial Arabic OCR. You need to have a dedicated OCR solution that you can train on your samples and specify your specific processing rules.



You can still use Tesseract, but in the form of your source code and tutorials to build your own solution yourself. To set up Tesseract for Arabic you can find this link helpful arabicocr.wordpress.com

0


source







All Articles