6

Is there any good free Hebrew OCR for Linux? Hocr doesn't count, as it's off such poor quality. I send computer generated text (not scanned, just saved as a gif) through it and it can't pick up anything.

image with Hebrew text

This seems like a fairly straightforward image to OCR. It doesn't look like what passes as a CAPTCHA these days, yet I've heard that OCRs are able to parse them. If so, all the more so this image. There are no curves, lines cutting through text, etc. Yet hocr can't OCR it.

If not, at least something trainable (I give it 500 gifs of letters, and then it will do OCR based on that).

  • 1
    Welcome to Software Recommendations! Could you provide an example image with expected result so we could vet other tools against it? – Angelo Fuchs Jul 17 '15 at 9:02
  • Maybe some finetuning was also needed. If you can give him some "advice", f.e. where to look for its lines. – user259412 Jul 18 '15 at 9:06
  • Did you try tesseract? There are also good tools to train it, for instance terese.sourceforge.net or more complex ones: succeed-project.eu/wiki/index.php/… – Nemo Oct 5 '15 at 17:11
  • @ocr, Did you find something usable over the past 3 years? – einpoklum Sep 1 '18 at 14:12
0

Yes, I use Tesseract. It provides language files for Hebrew language also. Following is an output of the package available on Debian GNU/Linux. This software is licensed under Apache License, Version 2.0

$ apt search tesseract-ocr-heb
Sorting... Done
Full Text Search... Done
tesseract-ocr-heb/oldstable 3.04.00-1 all
  tesseract-ocr language files for Hebrew

The core package is tesseract and language file for Hebrew is: tesseract-ocr-heb:

Package: tesseract-ocr-heb (1:4.00~git30-7274cfa-1)

tesseract-ocr language files for Hebrew

Tesseract is an open source Optical Character Recognition (OCR) Engine. It can be used directly, or (for programmers) using an API to extract printed text from images. This package contains the data needed for processing images in Hebrew language.

To use it, execute the following command:

tesseract -l heb /path/to/image.png /path/to/text

Here -l heb specifies the language, where image.png is the image containing Hebrew text. The output will be stored to text.txt file.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.