[sane-devel] Which scanner for doing good OCR under Linux ?

[Felix E. Klee]

On Wed, 31 Mar 2004 00:00:29 +0200 S=E9bastien Hinderer wrote:
> This scanner should be as fast as possible, and the scanned image should =
be
> of sufficient quality to allow an OCR to do a fine job with it.

You may want to try out a Bookeye scanner. In the university where I=20
study, there is one available in the library and I have tried it out=20
recently. After getting used to some of its quirks, I must say that it=20
seems to work pretty well. I haven't tried out OCR, though, and I don't=20
know if there are LINUX drivers for it. For more info, visit
http://www.imageware.de/wf_content.php/100.html .
I'm afraid, however, that these scanners are very expensive.
=20
> Apart from DPIs, which parameters can influence the quality of the text
> recognition ?

Scanning pages in thick books may be problematic because of the binding.
The Bookeye scanners have built in technology for binding correction.

By the way, in my university's library they also have a work place for
blind people that includes a scanner. I don't know which one that is,=20
though. But maybe a library near you has similar facilities.

Felix

PS: To contact me off list don't reply but send mail to "felix.klee" at
the domain "inka.de". Otherwise your email to me might get automatically
deleted!