Quicker than the Keyboard
By Joshua Hale, BuyerZone Staff
August 1st, 1999
Optical character recognition software can help push your paper problems away.
Whatever you do, don't let your staff of professional typists, data entry clerks, and
information specialists know you're learning about optical character recognition (OCR)
software. If they knew, they might be concerned by the fact that a scanner and some nifty
software can do their jobs in a fraction of the time at a fraction of the cost.
The value of a scanner to a business increases greatly when combined with OCR software.
When a document is scanned into a computer, the text in the document is readable, but not
editable. This is because the scanner captures the document as an image. OCR software expands
on the scanning process by reading individual characters and entering those characters
into an editing program such as Word. Documents scanned using OCR software can be updated,
edited, or filed for later use.
OCR is an example of an idea that has had to wait a little bit longer than the mainstream
acceptance of the personal computer before becoming a useful business tool. Many early
versions of different OCR software programs were plagued with errors that would negate
the time saved by scanning. Natalie Miller, product manager for Caere's OmniPage Pro OCR
software, believes that such programs have since caught up to, and exceeded, their human
counterparts. "Today there is no doubt that if you put a human typist next to an OCR engine," she
says, "the OCR engine will pull out the end result much faster and more accurately."
Using a clerk who can type seventy words per minute, it would take more than an hour to
transfer a 5,000 word contract to a computer, assuming the clerk typed error-free. With
an accuracy rate of ninety-nine percent, a scanner and OCR software would require just
minutes to transfer the same information. Wallace Schwab, president of Maurepas Services
Ltd., a Quebec City-based translation company, views his OCR software as the key to a quick
turnaround. "It's a productivity accelerator," he says.
Schwab uses OCR in conjunction with other translation programs, and notes that the increased
attention to detail in deciphering written text means his software, OmniPage Pro 9.0, can
handle unusual French and Spanish letters. "They've tried to make it as user friendly as
possible," he says, "where different accents, diacritical marks, aren't as cumbersome to
deal with."
Similarly, the latest versions of OCR software have been able to jump another early hurdle
- tables, charts, and other non-text marks on scanned documents. When Schwab received the
latest upgrade to OmniPage, he tested its ability to reproduce charts by scanning the economic
tables in The London Economist. "I really wanted to make the system squeak," he says. It
managed to recognize the small tables and eight-point font with only one error. "It screwed
up on one thing," recalls Schwab, "but when you're getting a less than one percent screwup
it becomes a joke."
One area that still trips up most OCR programs is paper and printing quality. A document
with many wrinkles, a coffee stain, and correction marks is probably going to be more prone
to OCR error than a clean sheet fresh off of the printer. Says Miller, "if you mangle the
page the recognition is not going to be as accurate."
Fred F. Ross, author of OCR with a Smile, a guide to OCR scanning, points out that some
programs handle quality issues differently. "One document may be a dot matrix document," says
Ross, "and there is definitely one software that will do better [with that type of document]
than the other."
A simple solution for addressing varied results is to run more than one OCR program against
a scanned document. Quickly comparing the errors among the three programs is an easy way
to see which software handles different types of scanning best. "For a small business owner
to go and buy two or three pieces of software is a very prudent decision," says Ross. "In
the long run they will save multiples on labor in form of editing."
He points out that for big OCR projects, a few errors per page can add up. "If you have
a 50-page document and scan it on three programs, you can cut down the error rate from
15 to seven errors per page and save 400 errors on the document that would take you two
or three hours to fix on the other end," he says.
With accuracy rates still reaching for perfection, manufacturers are moving towards more
experimental methods of reproducing the written word. A new type of software, known as
Voting OCR, enables the computer to decide the 'best' choice among different OCR programs,
and then automatically place confidence in that program for the chosen area.
Says Ross, "it will extract from the three pieces of software what it believes to be the
best situation, and that will cut your errors rate as much as eighty percent against using
one single piece of software." While Voting OCR eliminates most errors it is much more
expensive than typical programs and not recommended for a small business unless it specializes
in data entry.
Miller sees OCR development heading into aspects of artificial intelligence as technology
improves. "We're trying to get the software to look at the document as a whole and make
decisions based on the overall feel of the piece," she says. Ross says the next step is
to make "software that would analyze the entire sentence, and in the context of the sentence,
find the missing word."
Surprisingly, business computers, and not programming technologies, are currently holding
back entry into the next level. "I don't think the computers that we have today are powerful
enough to do this kind of recognition at the sentence or paragraph level," says Ross.
While OCR software isn't yet perfect, it is a proven time saver for businesses in the
process of converting paper-based documents into digital resources. If your business would
be improved by saving a room full of paper on a floppy disk, OCR merits your investigation.
Take care, though, when it comes time to pick your program.
The OCR food chain runs the gamut, from entry level to top notch. If you plan on scanning
the occasional memo, the stripped down OCR program that comes bundled with your scanner
will probably make do. In the same vein, if after reading this article you're ready to
dismiss your loyal staff of typists, you owe them the courtesy of learning which OCR program
will do the best job of replacing them.