I'm not sure if this is exactly answering your questions, it wasn't completely clean what you are trying to find out here... But, if I was doing a tunnel mapping I would still act as if the image was a flat image and just send a masked version of sections of the image to each projector. Just underst...