Part of the research I did in the last 3-4 years at the university I work for was based on spatial augmented reality (Spatial AR). The basic ideal of spatial AR is having the virtual augmentations appear directly over real objects, instead of using a screen or glass to superimpose them on the sight of the viewer, and the prevalent way to do that today is using projectors or projector-camera systems.
The focus of our research was at using low-cost tracking trying to create applications that make some practical sense (as in they are not only simple demos, they could eventually be useful) and study interaction based on mobile and portable projectors to try to make such programs usable at all.
We achieved some moderate success (defined by achieving the Best Paper Award at the most important conference in the area in Brazil - a shorter version of this paper) . I'd like to point out that I am just a collaborator and co-author of those works, Romero Tori, Eduardo Jacober, Cláudio Bueno and our students Lucas Trias, Newton Nakamato and Mark Hodgkin were also heavily involved.
Our next project was a game using spatial augmented reality, specifically tracked mobile projectors capable of delivering 3D augmentations nicely registered on the top of real objects . This project is on hold for now, because I am in a leave of absence from the Professor job to try to finish my PhD (in some private colleges and universities in Brazil you can still get a teaching/researching position before completing your PhD), and Eduardo went on to shift his career focus to a government IT career and being a professor just part-time.
Well, this will be completed some day, maybe with the help of some students. For now I'd just like to share some conceptual sketches. The nice thing about being an engineer/programmer that works with interactivity and games on occasion is that I get to put my drawing skills to some useful purpose.