[sane-devel] OCR software

[Wolfram Heider]

Am Dienstag, 16. Mai 2006 00:04 schrieb Olaf Meeuwissen:
> Greg Furmanek <greg at esdlc.com> writes:
> > Hi gang :-)
> >
> > I have been lurking on the list for quite some time.  There are a
> > lot of good info I got from you guys.  I am just wondering if you
> > have any recommendations when it comes to open source OCR libraries.
> > It would be nice if they were portable to win32 but not necessary.
> > I am mostly interested in the recognition algorithms and techniques.
>
> I am only aware of the following bits of free software.  Have not used
> any of them.
>
>   gocr  - (sorry no URL handy)
>   hocr  - http://hocr.berlios.de/
>   ocrad - http://www.gnu.org/software/ocrad/ocrad.html
>
> BTW, hocr-gtk says it's for Hebrew OCR.
>
> Hope this helps,
> --
> Olaf Meeuwissen
> FSF Associate Member #1962           sign up at http://member.fsf.org/
> GnuPG key: 30EF893A/2774 815B DE83 06C8 D733  6B5B 033C C857 30EF 893A
> Penguin's lib!       -- I hack, therefore I am --               LPIC-2

Having done extensive testing with open-sourrce-OCR-software I'm sad to say 
that this is one of the last real deficiencies of  open source against the 
Windows-world. None of them can seriously compete with commercial programs - 
either they are in their early stages or they are maintained any longer. 
So the only way to get reliable results under Linux is (especially with 
serif-characters) is to try to get running a Windows-program via WINE. 

Some in depth programming knowledge provided a further possibility may be to 
get a SDK for Linux from ABBYY (http://www.abbyy.de/sdk/?param=55304).

Wolfram Heider