imageimage-processingocrabbyy

Abbyy Finereader command line usage / python usage?

I have a project which needs to scan certain images with OCR. I know the software abbyy finereader does pretty well (also tried a trial version which works nearly perfectly for me) and now I'm wondering how to embed this software into python or another sripting language so that I can later simply run a command line script, e.g. python the_script.py image.jpg to take an image as parameter, execute abbyy finereader with it and return the recognised text...

I couldn't find a way on google, is there none!? I only found something like this: https://github.com/mikesname/python-ocrlab/blob/master/ocrsite/ocrlab/nodes/abbyy.py but couldn't work out what it really does...

Or which piece of abbyy software would I need to do this!? The Recognition server or the OCR engine or maybe something else?

Thanks a lot in advance

Solution

  • I have been working with ABBYY technology for over 15 years. I used to work for ABBYY, and now I work for WiseTREND - one of major ABBYY technology VARs and integrator in USA.

    FineReader, namely the retail boxed product, is designed for standalone interactive use by an operator. Command line and API automation is not available in that package. If your usage is small volume, you could use FineReader Corporate Edition as a simple black-box, set it up as a Hot Folder, and have your script drop images into that Input folder, wait for processing, and pickup from Output folder.

    For better automation and integration, a different product than FineReader needs to be used.

    First, decide if you would like to have local software or a cloud-based OCR.

    Cloud - ABBYY, OCR-IT, and a few other providers offer OCR via a Web request. Usually pay per page, and no software purchase required.

    Local - some software will run on your server and provide your scripts whit OCR functionality, as black-box or via direct API. Recognition Server is the next product to consider for this. Lastly, if you are developing a large application with the need to distribute it to numerous users, or need in-depth OCR control, then Engine SDK is the product to consider.

    If you would like to share more how you would like OCR to be integrated, and what volumes and functionality you need to achieve, we can try to narrow down even more.