Re: How good is Linux OCR?

Stefan Claas <stefan@mailchuck.com> wrote:

Hi all,
 
a quick test with my 'atze' encoder/decoder and Microsoft OCR
gave no errors with this blurred text image. Can you archive
the same error free results under Linux and if, what program
do you use?
 
https://jmp.sh/MUJtkvAr
 
I used with my encoder 2000 bytes, converted to 8000 chars,
per A4 page. This is not as much as with stacked QR-codes,
but readable by humans and therefore ideally suited for
encrypted binary data.
 
Microsoft OCR is pretty awesome.

There is no such thing as "Linux OCR" (as in an OCR provided by Linux).

Tesseract is probably the most common one used on Linux, because it
(Tesseract) is open source and its source will compile with Linux C
compilers.

But that hardly makes it "Linux OCR" (not in the way that one would
mean with "Microsoft OCR" -- as in provided by Microsoft).

Note that Tesseract will (I think) compile for windows too, so if you
wanted to know "how well tesseract worked" you could just install the
windows version and see for yourself.