the usual way is to either directly process pdfs in perl with the help of modules (search cpan), but that is fairly involved. so sometimes one would use a utility like xpdf or such to convert pdfs to text or html, and then find a way to extract the required information. it seems you're trying the opposite to the usual tactic, and instead of parsing html (much easier) as per your first post, you've gone to pdf and again ask for code to fall from the sky without showing any programming effort yourself. if you just want the job done and you have no clue about perl or care about programming, then just post it up on some other forum that's geared towards providing free tools/code/advice to support your efforts. this is a perl forum for people interested in perl programming...or at the very least people interested in and actively learning technology. i don't want to disuade you from your efforts, but i do question whether this is the right forum/audience for you.
in reply to Extracting caption of a image from PDF file
the hardest line to type correctly is: stty erase ^H