Several investigators have tried IR cameras, including the SPARWARS group.
The video is posted on several siites, but I don't have the URLs at hand.
The video is of codeposition on a nickel screen and is fascinating to watch,
but I'm not sure what it tells one. The ordinary CMOS and CCD sensors are IR
sensitive, but to get high resolution images you need IR optics and it
begins to get pricey. The cell has to be built so the cathode is close to a
window, for water is IR absording. As Jed points out, such notions become
projects in themselves.
Many years ago I tried to think of desktop CF toys that would not be
spectacular but very persistant, without success. Sitrling engine toys that
will run off the heat of your hand are in the right direction.
Mike Carrell