The Sorcerer's Workshop

Do you know what the acronym OCR stands for?

Is it:

Online Computer Reviews
Oil Conservation Route
Owl Care Room
Outer Core Reactor
Onion Cutting Routine
Optical Character Recognition

If you chose ‘Optical Character Recognition’ you can read the rest of this article. If you chose ‘Owl Care Room’ or one of the others, you can still read this article but maybe Web design is not for you.

OCR – Optical Character Recognition – Software designed to accurately convert scanned text documents into computer editable text. OCR technology use algorithms to translate combinations of dots in a bitmap into a recognized character.

OCR is a technology that has been around for years. In the beginning, the accuracy claims by software companies was somewhat overstated (i.e. very exaggerated). Users often had to spend hours rechecking converted documents because the OCR software saw an ‘i’ and turned it into a ‘1’. Alternately, it would convert italicized text it into a something only a CIA operative with a decoder ring could decipher. Fortunately, most of these software glitches have now been resolved. At least that’s what the makers of OCR technology would have us believe. Most claim an accuracy rate of 99%.

A Real World Project

Recently I received permission to place a sample book chapter on our iBoost Web site. Peachpit Press graciously gave us the OK to put Chapter 10 of ‘HTML 4 for the World Wide Web’ by Elizabeth Castro online. Its an excellent book, and I highly recommend it.

Screenshot of TypeReader 6

Unfortunately for us, they could only give us the book in Framemaker format. Not having the program, I needed an alternative. My choice was to OCR a hard copy of the chapter. But OCR software is not cheap. OmniPage Pro 10 runs about $500. TypeReader Pro 6.0 from Expervision costs almost $300. Both claim to be the best, boasting 99 percent accuracy, the power to recognize foreign languages, export the results to HTML and make a great cup of coffee all the while I sit back and try to decipher Dennis Miller's latest obscure reference on Monday Night Football.

After very, very careful review of both programs, examining all the pluses and minuses, and expending hours of serious meditation (i.e. playing Gameboy) I chose TypeReader 6.0. The deciding factor was Expervision’s accuracy claims, its support of over 2600 fonts and most importantly the use of a free 30-day trial version. I downloaded the 15 meg program and installed it without a hitch.

(1) (2)