kusumoto@chsun1.uchicago.edu (Bob Kusumoto) (01/31/91)
Calera makes WordScan (and WordScan Plus, WordScan/AT, WordScan/MCA), which
is a nice program (we picked it up). It got a decent review in a past
Infoworld article (did fairly well in accuracy, not as well as Omnipage
though) and got an editors choice from PC Week (against HP's AccuScan and
some other lame OCR program). It's a nice program that works in Windows 3.0
and does a decent job of doing OCR. It's my opinion that it handles cleaner
documents swimmingly but doing something like scanning in the sports page
stats or handwriting is tough. It's slow (according to Infoworld), but
it converts the OCR into the widest range of text and graphics formats
(definate plus) and does deferred processing (so you can do those complex
scanning of 20 pages with your HP Scanjet Plus with an automatic document
feeder, without having to be around). The AT and MCA versions is a hardware
card that speeds up processing but costs a hell of a lot more.
Caere makes Omnipage, which is another great OCR program (if you can afford
it). Infoworld reviewed the Windows 386 2.11 version and it was the most
accurate program against the others. It also commented that it was relatively
speedy. They've since come out with a Windows 3.0 version (which supposedly
works in 32-bit mode or something or other) and repackaged it somewhat into
Omnipage and Omnipage Pro (competes better with WordScan and WordScan plus).
I just wished that it was out when I bought WordScan Plus, since it is
faster and it can deal with handwriting a bit better (WordScan can be
pretty awful with handwriting, but in any case, it has to be pretty neat
handwriting to begin with, so take this with a grain of salt). Caere seems
to be OEM versions of their product to hand scanner makers but I'd get
a demo from a sales rep from your local store and bring samples of what
you intend to scan in a lot so you can see which OCR program will suit
your needs. It seems that this is pretty sound advice since some programs
tend to deal with certain types of documents better than others (reference
the Infoworld article, they did samples from a double spaced, monotype
document, a magazine article, a newspaper article, and some other sample
for their speed and accuracy tests), that way you have the best program
and scanner for the types of things you're going to be dealing with.
Bob
--
Bob Kusumoto | I just come from the land of
Internet: kusumoto@chsun1.uchicago.edu | the sun/ from a war that must
Bitnet: kusumoto%chsun1@uchicago[.bitnet] | be won in the name of truth.
UUCP: ...!{oddjob,gargoyle}!chsun1!kusumoto | - New Order, "Love Vigilantes"