Re: scanning tables, line art into FrameMaker

Subject: Re: scanning tables, line art into FrameMaker
From: "Dan S. Azlin" <dazlin -at- SHORE -dot- NET>
Date: Mon, 5 Feb 1996 16:29:46 -0500

On Thu, 1 Feb 1996, Mark I Halpern wrote:

> Colleagues,

> I am responsible for recommending to my company a scanner/OCR product that
> will enable us to turn hard copy, as directly as possible, into FrameMaker
> files. I am finding that the low-end products, such as those reviewed in
> PC Magazine in October '95, seem not to be able to do this directly; they
> seem to involve scanning, applying OCR software to convert the scanned
> image to an ASCII (or Word) file, then importing that file into
> FrameMaker.

> So my first question is, does anyone know of a product that will scan a
> page of text directly into FrameMaker? (Logitech tells me that they
> can do this with any wordprocessor that is TWAIN-compliant; Adobe's
> Tech Support desk wasn't able to tell me if FrameMaker is
> TWAIN-compliant; can anyone tell me?)
Check out the "Document Management" article in Computer Shopper's January
issue (also available at for some candidates.
What you are asking for is a pretty tall order in any environment.
Scanning in any document and translating the visual image into editable
text is the function of an OCR engine. Usually the text is translated
into ASCII text, which can be translated into a word procesing file
format with relative ease. FrameMaker can import several standard word
processing formats which can make the process easier. However, if you
expect the document to have embedded formatting, paragraph tags, etc.
Forget it. There isn't anything out there but a tech writer who can do
that job :).

Keep in mind that even commercial RIP engines like the DocuTech printing
system converts scans into bitmapped tiff files prior to printing. Its
the state of the art.

> Secondly, we would like the product to be able to scan tables and line art
> in "intelligently"; that is, to scan them in not simply as graphics, but
> to turn them into editable constructs, just as it turns printed characters
> into editable text.
Again, this is beyond the veil. Understand that the scanned image is a
bitmap. Regardless of the format you save it in, it remained a bitmapped
image. There are tools that will generate an autotrace of the image and
give you a polyline vector representation of the image. But this usually
has to be cleaned up to be useful. Most of the major graphics packages
have this ability (Corel Draw, Adobe Illustrator, etc.) The only tool I
know of than turns a scan directly into a vector image is an add on for
AutoCAD that works with a plotter and a scanning head attachment.

Sorry I couldn't offer something better.

Dan Azlin ** WORD ENGINEERS, Technical Writing & Publishing **
Ph/Fax: 508-921-8908 18 School Street
Internet: dazlin -at- shore -dot- net Beverly, MA 01915-4851

Previous by Author: Looking for Lotus writers...
Next by Author: Re: Image formats for tech. comms.
Previous by Thread: scanning tables, line art into FrameMaker
Next by Thread: User, User's, Users' Guide

What this post helpful? Share it with friends and colleagues:

Sponsored Ads

Sponsored Ads