Converting DJVU with bad OCR

[Ridcully]

I have a DJVU file that I want to read on Kindle, which has high quality scans with little noise, all properly oriented etc. However, when I try to convert it to any other format (including PDF) with Calibre, I end up with the underlying OCR text, which is horrible - no paragraph breaks, probably never checked for spelling, any bold/italic text is completely garbled... How do I get rid of this and get the scanned images instead?

[doubleshuffle]

I have no idea how to do that in Calibre; I guess you will have to run your file through a proper OCR program like ABBYY Finereader (or one of the free OCR applications based on Tesseract). I don't know which of them process djvu, though; perhaps you will have to save the individual pages as images first.

Anyway, no matter how well your OCR turns out, you will always have to manually proofread the result.

[Divingduck]

In addition, if you have no OCR you can print the DJVU-file alternatively to a PDF-Printer (included in windows) and open the PDF in MS-Word 2016 (or version 2013). The included OCR in Word did a surprisingly good Job. It's not perfect but in combination with spell checker and S&R you can do a decent job too.

[Ridcully]

Can't I just remove the OCR text inside the DJVU file somehow? I'd prefer to have the images in PDF format on Kindle, the letters are clear and the pages are narrow, and I wouldn't have to deal with the whole OCR thing.

[doubleshuffle]

I've just done a quick web search for "djvu to pdf conversion". It seems there are tons of free online tools available; why don't you try one of those?

[Tex2002ans]

If I am understanding correctly, you want your DJVU (scanned images with hidden OCR) to be turned into a PDF (scanned images only).

There are multiple ways to handle this.

If you aren't afraid of the commandline, you could use ddjvu to go from DJVU -> PDF:

http://djvu.sourceforge.net/doc/index.html

For example, here is one usage of ddjvu off of Stack Exchange:

https://superuser.com/questions/1005.../596167#596167

[Ridcully]

Recently had the same problem again (and forgot the solution), but this time around I found a GUI program before I remembered this thread. Apologies for bumping this long deceased thread, hopefully this helps someone else who googles for this.

In case anyone doesn't want to use command line, a non-command line freeware program does just what I wanted - it converted the DJVU to PDF, preserving the image layer and TOC but discarding the OCR layer: STDU Converter. It has lots of options but I just hit Convert and it did what I wanted, and gave a fairly small output file too.