dplatt@coherent.com (Dave Platt) (12/07/89)
A couple of weeks ago, I posted an article asking if anyone had successfully
used a Mac with a fax-modem and an OCR package to receive and interpret
facsimile images of printed text. I received several responses, from
Ian Feldman, Mark Anbinder, Will Ray, and Jim Morris among others.
To sum up the answers: this combination isn't very workable in practice.
Most OCR packages really need the 300 dpi resolution of a good scanner;
the 200 dpi resolution of fax images isn't really sufficient for accurate
character recognition. Also, the fax-sending process tends to introduce
a significant number of errors... due to line noise, dust and dirt on
the page or on the sending fax machine's scanner, and so forth. To make
matters worse, pages are often fed into a fax-machine slightly off-true
(crooked); this doesn't matter much to the human eye, but can really
confuse OCR software.
The overall conclusion is that text processed in this way (printed onto
hardcopy at 300 DPI, fax'ed to a faxmodem-equipped Mac, and OCR-
interpreted) would have a significant number of errors in it, and would
require a substantial amount of hand-editing. This isn't good enough
for our purposes... we are trying to get our software-support people
out of the text-entry/editing business entirely.
Several people suggested that the process might work if the image was
sent via a faxmodem-equipped Mac, rather than by a standard fax machine.
Unfortunately, this option probably isn't acceptable to the customers
we're supporting... we don't want to require them to install this much
extra hardware, or dedicate somebody to becoming proficient at its use.
We'll probably use another method... perhaps having the customers install
a standard dialup modem on one of their workstations (Suns, mostly);
we can then dial up and download the text-files via ZMODEM or something
similar.
Thanks to all who responded to my plea for information!
--
Dave Platt VOICE: (415) 493-8805
UUCP: ...!{ames,apple,uunet}!coherent!dplatt DOMAIN: dplatt@coherent.com
INTERNET: coherent!dplatt@ames.arpa, ...@uunet.uu.net
USNAIL: Coherent Thought Inc. 3350 West Bayshore #205 Palo Alto CA 94303bs1k+@andrew.cmu.edu (Bryan Todd Schmersal) (12/07/89)
Another reason that Faxmodems and OCR's really won't work together is that nothing is standardized. Many times, cover pages are hand written or you receive forms that have hand writing on them. This was the reason we, when we were working on our paperless fax system here at CMU for our software engineering class, didn't even consider trying to encorporate OCR in our project. SMURF ----------------------------------------------------------------------------- CCCCC M M U U | Bryan T. Schmersal ==> SMURF bs1k+@andrew.cmu.edu C MM MM U U | Carnegie Mellon University (412)268-8579 C M M M U U | Pittsburgh, Pennsylvania 15213 POST C M M U U | "I'd rather be a survivor at Carnegie NO CCCCC M M UUUUU | Mellon than a hero at Pitt." BILLS -----------------------------------------------------------------------------