hi David and Others
One of the issues with Omnipage Pro 9.0 is that, by default, it
uses TWAIN drivers, instead of the ISIS drivers used by version
8.0. What this means in practice is that it will almost certainly
take much longer to scan a page. I'm using an HP 5P scanner, and
using the ISIS driver I can scan a page in 8 seconds. Using the
TWAIN driver, it takes about 23. It is possible to use Omnipage 9
with an ISIS driver, but it takes a bit of configuring. The TWAIN
standard is (or, rather, has become) becoming the industry
standard, for reasons that have as much to do with politics as
anything else, and if you're only going to be scanning small
quantities of material the increase in speed may not be as much
of a problem as it is for those of us who scan lots of
full-length books. (There is a big difference between 3 hours and
9).
Another thing I've found is that Omnipage 9 sometimes completely
ignores a page of ordinary text, interpreting it as a graphic.
These same pages are recognised correctly by Omnipage 8, and the
problem seems to be related to the autozoning algorithms used by
Omnipage 9. That is, Omnipage looks at the page and figures out
whether there are pictures on it or not; if it thinks that there
are pictures (graphics), then it won't OCR that part of the page.
For some reason, Omnipage 9 sometimes thinks that there are
pictures when, in fact, there are not. Using the zone drawing
tools that are part of Omnipage 9 it is possible to force
Omnipage to treat a segment of a page (or the whole page) as
text, but as far as I am aware it is impossible to do this with a
screen reader because the page image is a graphic before it is
OCR'd, and so cannot be accessed with speech or braille.
I have been trying for several weeks now to get Caere Technical
Support to acknowledge and investigate this problem, and I've
sent them files illustrating the problem. However, there seem to
be quite a few process issues that have prevented their response,
and I have taken these up with Caere senior management. However,
the situation so far is still unresolved. I therefore don't know
if the problem is solvable within the context of the
currently-released version.
Again, unless you are scanning lots of material, you may not
encounter this problem, but I have found that at least one page
in every book that I have scanned with Omnipage 9 is ignored. For
me this is a sufficiently serious problem to have led me to look
at other OCR products, and I'm in the process of investigating
Typereader, Finereader, Cuneiform OCR, Charactereyes
Professional, and Readiris (I already use Textbridge as well as
Omnipage).
As for the specifics of using Omnipage 9 with screen readers,
I've found that it works well with Jaws and braille (presumably
also with speech), except that there seem to be some problems
labelling the graphics on the zone toolbar (but you probably
won't want to do this anyway).
Cheers
Bruce Maguire
[EMAIL PROTECTED]
On Thu, 28 Jan 1999 00:49:37 -0800, you wrote:
>Hello listers:
>
>I just recievd a copy of OmniPage Pro V9.0 and wanted to know what issues
>there are with the program. this is my first scanner with OCR. I have
>successfully scanned a couple of files, but I wanted to know what to
>watch for.
>
>any inco would help.
>
>
>From: David Edick<[EMAIL PROTECTED]>, Engaged to [EMAIL PROTECTED]
>see the new juno list at: http://www.onelist.com/subscribe.cgi/junolist
>my Email lists can be found on my WebPage:
>http://www.angelfire.com/wa/mmdlep
>Phone: (360) 876-9712 Jesus Is, Lord! of my life,
>
>___________________________________________________________________
>You don't need to buy Internet access to use free Internet e-mail.
>Get completely free e-mail from Juno at http://www.juno.com/getjuno.html
>or call Juno at (800) 654-JUNO [654-5866]
>-
>Visit the jfw ml web page: http://yoyo.cc.monash.edu.au/~nallan/jfw
-
Visit the jfw ml web page: http://yoyo.cc.monash.edu.au/~nallan/jfw
