Re: [hylafax-users] OCR suggestions

["Graham Chiu" <compkarori@xxxxxxxxx>]

Well, I'm not ready yet to try ocrad :)

My script now rotates the incoming fax to 180 degrees if it can't
reasonably decode at 0 deg.  If that fails, it then rotates the
orginal to 90 deg, and if that fails it tries 270 deg, and then gives
up!

By reasonable, I mean I randomly sample words from the ocr'd text, and
check them against the 1000 most common english words, and see what
score I get.

Tesseract seems to work quite well but seems to be font sensitive.
Eg. it wasn't able to decode a large non Times-Roman font, but was
able to cope with a smaller times-roman.

On 1/31/07, Uwe Dippel <udippel@xxxxxxxxx> wrote:
> Graham Chiu wrote:
>
> > Big test today failed though .. as the faxes came in upside down and
> > so ocr failed :(
> >
> > Must be a way to detect upside down images and invert them again ??
>
> (Feel like selling you an old idea, since you never take the bait :( )
> We do this with ocrad, it permits to rotate the image. So we try to
> extract the e-mail address, do some checking. If it isn't, rotate.
> The whole runs in a loop written in shell script.
>
> Uwe


--
Graham Chiu
http://www.synapsedirect.com
Synapse-EMR - innovative electronic medical records system

____________________ HylaFAX(tm) Users Mailing List _______________________
 To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
 *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*