Is there a combination of utlities (i.e. ImageMagick,
tumble, etc.) that
can automate this instead of having to recreate each PDF from scratch?
Your two best friends:
1. Ghostscript. It'll take PDF's as input and produce anything you want
(postscript, pnm, png, etc.) as output.
2. The CPAN module PDF::API2. For truly structured documents where you
want to manipulate the innards this is heaven. Very nice object-oriented
interface.
I suspect Ghostscript will be of more immediate use because it sounds
like each PDF page has very little structure and is simply a big bitmap
image.
Tim.