OCR software continues to improve. When BYTE reviewed OCR software three years ago, the character-recognition accuracy of some packages was as low as 60 percent. Now these programs can typically recognize 96 percent--or higher--of a scanned document correctly while integrating with your word processing applications and preserving such format information as multicolumn document layouts; font attributes, such as boldface and italics; headers and footers; and tables.
The latest example of this trend is the Windows version of TextBridge Professional Edition 3.0 ($349; a Power Mac version is slated for release in June). The preproduction version BYTE previewed installs itself into Windows word processors, including Microsoft Word, WordPer
fect, and Lotus Development's forthcoming WordPro (formerly known as Ami Pro), so you can recognize and proof documents within your word processor. If your word processor supports multicolumn layouts, you can take advantage of TextBridge's ability to preserve multicolumn layouts to obtain further time savings. TextBridge is also intelligent; it can recognize a header or footer in a scanned document and maintain that designation when it converts the document to a word processor that supports headers and footers.
The accuracy you see depends on such variables as the quality of the faxes or documents you scan and the point size of the fonts used. For instance, when I asked TextBridge to recognize a faxed press release containing 12-point skewed text (see the screen), its character-recognition accuracy was 99.4 percent. The program also has an interactive training component that lets you teach TextBridge to recognize unusual words, such as product names, to further increase accuracy. This heightened accura
cy, together with the program's integration with other applications, indicates that TextBridge Professional Edition 3.0 for Windows will prove to be a worthy companion for your word processor.
BYTE Digest editors every month analyze and evaluate the best articles from Information Week, EE Times, Dr. Dobb's Journal, Network Computing, Sys Admin,
and dozens of other CMP publications—bringing
you critical news and information about wireless communication,
computer security, software development, embedded systems,
and more!
Preview: Xerox's Solid Paperless-Office Bridge
