Full layout and text region classification capability of Cloud OCR SKD

  • 2.5K Views
  • Last Post 28 July 2015
  • Topic Is Solved
kungfujam posted this 18 July 2015

I'm trying to recognise semantic information about text regions e.g. headers, footers, page numbers, captions etc. I cannot find any information about how to achieve this in the documentation. Please can you provide a python code example?

I am using the supplied python example here: http://ocrsdk.com/documentation/code-samples/. In order to get this Python code to work, I had to change line 90 of AbbyyOnlineSdk.py from

return {"Authorization": "Basic %s" % base64.encodestring( ...

to

return {"Authorization": "Basic %s" % base64.b64encode( ...

I'm currently just executing this call:

python process.py infile.png outfile.xml -l English -xml

Edit


This issue is now fixed on GitHub: https://github.com/abbyysdk/ocrsdk.com/issues/27

Oksana Serdyuk posted this 28 July 2015

This feature is not supported in ABBYY Cloud OCR SDK. But you can try to use another our SDK for your needs, namely ABBYY FineReader Engine. You can access the logical structure of a document formatting attributes, including headers, footers, page numbers, fonts, styles etc., via the IFRDocument::DocumentStructure and IFRPage::PageStructure properties. Please find more information in Help → Guided Tour → Advanced Techniques → Working with the Logical Structure of a Document.

For trial version of FineReader Engine, please fill the form at our site.

  • Liked by
  • kungfujam
Close