OCR software
Mie-Yun Lee, Editorial Director, BuyerZone.com
August 31, 1999
While happiness may be playing with a puppy, it certainly isn't having to type in a
document for which there is no file nor is it having to thumb through a stack of old
invoices in search of the right one. A scanner can help you avoid being completely unhappy,
though, when faced with these projects.
Under normal circumstances, a scanned image can't be changed once it's in electronic
form. The scanner reads a page as one whole picture, not as individual letters, words,
and paragraphs. Optical character recognition (OCR) software, true to its name, recognizes
each individual character of scanned text, then transfers them into an editable environment
such as a word processing or spreadsheet program.
Business applications for OCR software are plentiful. Turn your old files in the closet
into a searchable digital filing cabinet. Recreate invoices online without spending hours
tinkering with complex design software. Create spreadsheets of hard-copy sales data from
multiple locations in minutes instead of days. Turn a fax into an instant memo, letter,
or email.
While OCR programs have been around for about 10 years, their performance hasn't been
stellar. Early versions were fast, but the time required to correct misread characters
and other errors ended up negating the time saved, and using them for anything except
straight text was problematic. Most of today's high-end versions now boast accuracy rates
of 99 percent, and can handle different fonts, tables, and other formatting nuances with
relative ease.
Not all OCR software is equal, of course. Some programs deal with certain document formats
better than others--like dot-matrix printed pages, for example, or columnar newsprint
or table-heavy reports. If you expect to use OCR software as a regular part of your everyday
business, it may make sense to buy more than one OCR program and use each for different
scanning situations depending on the quality of the output.
Pricewise, software falls at both ends of the spectrum. At the low end, bare bones applications
commonly bundled with scanner purchases can handle OCR tasks, but are generally slower,
less accurate, and more prone to formatting trouble than the market leaders. The same
can be said for cheaper stand-alone products that sell in the $75-$150 range. But if
you can deal with a few errors and don't want to scan heavily formatted pages, these
bite-sized products can work just fine.
However, if you plan to use OCR software as a serious business tool, you'll want to
lean more to the higher end products that can be purchased online or at a computer store.
The serious contenders of the more advanced products include Caere OmniPage, Xerox TextBridge,
Expervision Typereader. Expect to pay between $300-$500 for a package.
Upgrades are common as the technology improves, but the next big jump for OCR (word
and context recognition using artificial intelligence) will have to wait until more powerful
PCs hit the business market. In other words, it 's a good time to get in now when you
probably won't have to buy an upgrade for a couple of years. Look at it as a chance to
get a head start on the paper trail.
Quick tips
Look to the bundle. If you're shopping for
a scanner and don't want to drop the big bucks for a stand alone OCR title, get
a stripped down version bundled with the new scanner and upgrade on the cheap.

Use competitive upgrades. Buying more than one program? Tell the seller you'
re using a competitor's product and considering switching, and you'll likely get
their package at a much lower cost.

Walk before you run. If you're in the market for a scanner and an OCR application,
check out the software titles first. Not all software supports all scanners, so check
first for compatibility. |