James,

  This is true.

  http://access.adobe.com/AccessGuide.pdf describes at least one path that can
be used to turn scanned documents into accessible PDF. However, any time the
software has to guess at the author's document structure, there will be
mistakes.

        Loretta

> I think that so long as document creators can produce a pdf file that is
> a graphic only without any OCR, then we will always have documents that
> aren't accessible.  Converting a graphic to characters needs to be an
> integral and embedded part of the process.
>
> Regards,
>
> James
>
> James Bailey
> Adaptive Technology Access Adviser, University of Oregon
> Office: 541-346-1076
> Adaptive Lab: 541-346-1927
> [log in to unmask]
> http://adaptive-tech.uoregon.edu
>