[vox-tech] Does anyone use Clara OCR?

[Doctorcam]

* Henry House (hajhouse at houseag.com) wrote:
> På tisdag, 28 september 2004, skrev Jan W:
> > I tried it out.  Wasn't all that great.  That was maybe 2 years ago. 
> > Haven't checked it out since then.  It had some sort of learning
> > function that had to be 'trained'.  Really wasn't viable for our OCR
> > needs, so I haven't looked at it for awhile.  You also might look at
> > j/gocr at:
> 
> I usually use gocr for scanned documents. It is reasonably sucessful if the
> resolution is high and the letters are crisp (inkjet printouts are
> problematic). But it does not work at all on low-resolution faxes. Hence my
> interest in a trainable OCR program.

I found Clara rather complicated to use, and the documentation was not
very helpful.  I abandoned it, mostly because I needed something I
could put to work right away.  With a Mac laptop, that was
easy. However, the idea of a trainable OCR still appeals to me, even
if (as I found) the training period takes longer than you might like.
It would work, but the first couple of passes were often such
gibberish it was hard to determine what it was working with -
correcting it was sometimes hit-and-miss.  Even the one I used on the
Mac would insert spurious entries because the scanner picked up a
speck of dust on the glass.

Sorry - I'm sure this is not much help.

Apt-cache show gives the following:

 Clara OCR is intended for large scale digitalization
 projects. It features a powerful GUI and a web interface
 for cooperative digitalization of books.

I see that the latest Debian version is 0.9.8, so it is probably a lot
better than when I worked with it.  I don't remember the version
number being that high.

Cheers

Cam



-- 
Cam Ellison  Ph.D.  R.Psych. #01417

Cam Ellison & Associates Ltd.
Management Psychology

RR 22    3446 Beach Avenue
Roberts Creek  BC  V0N 2W2

Phone: 604-885-4806
Fax:   604-885-4809