How to make OCR Engine read pdf documents better?

[galymbek]

Hey guys, I have a trouble reading pdf docs (scans of some invoices). I want to read them with “Read PDF with OCR”, then get the needed values, and then insert these values into some application. Those scans have not that poor quality to read them.

However, I can’t get these values, because the quality of extraction is bad. I use Tesseract OCR (I’ve read that it is a “usual” Google OCR).

What can you suggest to do?

Thanks in advance!

[huhuhug (Hung Nguyen)]

I refer you use Omni page OCR. My exp, It’s the best ORC now!

[galymbek]

No, I think ABBYY Cloud OCR is the best, although it has some limited number of pages. But it is much better than any other OCRs

[huhuhug (Hung Nguyen)]

Omnipage is free of cost.

[galymbek]

Try to make some poor quality scan version of invoice (pdf), then you will see the difference and you will understand that it is better to create new emails to register in ABBYY (for free) rather than use Omnipage

[huhuhug (Hung Nguyen)]

In my case, I convert one poor quality scan file with 2 OCRs and Omnipage better than ABBYY. May be dependent on some condition.