OCR

TextBridge Professional Edition 3.0 (November 1995)

Bridge Spans Scans

BY: Hailey Lynne McKeefry

My scanner came with a basic utility that let me stumble through document scanning. TextBridge Professional Edition 3.0, with its easy-to-use interface and strong OCR capabilities, makes that same journey a virtual walk in the park.

The program's interface is simple and clean-cut. Eight control-bar buttons provide access to many of TextBridge's features, including Start/Continue Processing, Stop Processing and Cancel Current Page. From this control bar, I could bring information into the program directly from a file or from a scanner. I could also specify whether I wanted to preview the scan before the OCR, defer OCR or train the OCR.

With the Instant Access OCR feature, you can access TextBridge from within any of about a dozen applications. Among my array of applications, TextBridge provided this support for Sidekick, Microsoft Word 6.0 and Microsoft Mail. By using Instant Access, I was able to "register" my most-often-used apps so that the TextBridge OCR command was added to each one's File menu. By keeping the Instant Access feature running but minimized, the OCR command stays active on the File menus so you can launch TextBridge quickly without leaving your current work.

Despite some of its rather sophisticated features, TextBridge simplifies document scanning. Place the page in your scanner, click on the GO button and the software activates the scanner, acquires the image and reads the document. At the end of the process, the software prompts for additional pages. You can save the output in any of 30 formats, including Microsoft Word 2.0 or 6.0; WordPerfect 4.2, 5.1, 6.0 or 6.1; Lotus Ami Pro 2.0 or 3.0; Microsoft Excel 3.0 or 4.0; and PostScript. (I saved the test documents as Word 6.0 .RTF files.)

TextBridge's preview allows you to create text zones on the page and choose only certain document parts for OCR processing. With the Image Zone tool, you indicate which scanned-image parts are graphics that should be saved without OCR processing. Doing OCR training on the document--defining consistently misinterpreted text--improved the already acceptable accuracy substantially and was no more time-consuming than spell checking a document. I was also able to rotate misscanned documents so that they were positioned correctly for OCR.

The included TextBridge Proofreader lets you perform post-recognition proofreading from within Word or WordPerfect. These proofreaders are more powerful than the OCR training. Suspect words are color coded, based on the program's confidence in its best guess: green for high, dark yellow for medium and red for low. The Proofreader lets you choose to Stop At or Accept All words in each confidence category. As with a spell checker, there's a Replace All button for those times when a term or phrase is consistently misrecognized. Show Image displays a bitmap image of incorrect words in their context.

Document Recomposition is, perhaps, TextBridge's most powerful feature. There are two recomposition modes: text-only, for documents with tables that should be output as cell tables; and text and automatically detected graphics mode, which outputs the page with original column and art layout intact. Very complicated layouts confound this feature, so it may be necessary to manually zone the text and images.

TextBridge Professional is a sophisticated OCR application that is probably overkill for those who need to scan only occasionally. But the program's straightforward interface belies its powerful capabilities.

--Info File--
TextBridge Professional Edition 3.0
Price:
$249 (street); upgrade, $149
In Brief: TextBridge is a sophisticated yet easy-to-use OCR package.
Disk Space: 13MB
System Resources: 1%
RAM: 8MB
Xerox Desktop Document Systems
800-248-6550 x3, 508-977-2000

WordScan Plus 4.0 (July 1995)

see Pagekeeper 2.0 (Document Management)