, actually... the cameras expect to read things the way they wrote them, so as long as you force your converted file to conform with this (which means all EXIF fields, thumbnail offset, image data offset, etc.) it will display the image just fine, instead of that question mark for unidentified images. The usefulness of it is another thing. Your zoom levels are quite limited, and if you (for example) convert a 8.5"x11" PDF page into a huge 1200dpi image, and paste it... you can only zoom in so far, and the usability / readability is a real PITA. Don't take my word for it, here's a way to try it:
I'll suggest grabbing XnView
by Pierre E. Gougelet... it's freeware. Grab the full
version, as it has all the plugins pre-installed and ready to go.
Fire it up, go to any image that you want to put on your camera in the Browser, and double-click it, do Ctrl-A, Ctrl-C (Select all & copy, or use the mouse and menus) Go to a disposable image that your camera took, and Ctrl-V, Ctrl-S (paste & save)... now just move it to your SD card and put it in the camera.
I sure you want a programmatic way of automating it, so you could say 'convert4camera some.pdf myimages*.jpg' - well you've got a lot ahead as I don't think any work has been done like this. But as you can see, it is
If you decide to go this route, good luck!