4

I have a fair number of formal linguistics papers (generative syntax, semantics) and I am looking for a good OCR software to digitize them as searchable PDFs. Unfortunately, the programs I have tried work well for pure text, but choke on IPA, logical symbols and any type of structural tree. Glossed examples with diacritics and indexes also fare poorly.

Can anyone suggest a scan + OCR software for Windows, Linux or Android that works well in the specific domain of linguistics?

Roberto Zamparelli

University of Trento, Italy

4
  • Could you add a list of the software you have tried and found unsuitable for your purpose? Does any of the programs you tried have the option of adding your own custom recognition language by defining its characters set? If the OCR software has such an option, you can add IPA by listing its characters.
    – Yellow Sky
    Commented Mar 26 at 14:38
  • OCR quality also depends on the quality of the original. I've had a lot of trouble but now there are new programs like Nanonets. Have you tried the new programs?
    – Lambie
    Commented Mar 27 at 15:50
  • Thanks. I have tried android-based OCR like Cam Scanner, and Acrobat Porfessional (admittedly, an old version). I think the problems cannot be easily solved by defining new characters (logical symbols in trees are an especially difficult problem), but I agree that they it could cover some distance. I have never heard of Nanonets (it should like an ANN-based software). I can check it out, thanks! Commented Apr 3 at 14:13
  • ABBYY Finereader is able to add new symbols to the list of the symbols it recognizes and also in it you create your custom recognition languages by choosing a language it's already able to recognize and adding more characters to the characters set of that language. I was able to OCR Persian texts with it although Persian is not on its languages list, I added the specific Persian letters to the Arabic char. set which is already there. Also, it has an option to edit the recognized text in the in-built text editor. Try it, I do recommend it.
    – Yellow Sky
    Commented Apr 3 at 22:53

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.