You might have seen various videos on the Web, where building come to life and seem to produce 3D art, like this:
Or maybe you have even seen the same effect done with cars, like here:
Well, the process of obtaining such a 3D projection that is mapped to a complex surface is in theory not very hard… It’s a lot (LOT) of work, but if you know what you are doing it is doable. It’s pretty much similar with what is presented in this video:
Now the main difference between this previous video and what we are talking about is that we don’t have an initial photograph of the scene. Instead, we would work a bit back-to-front for the implementation. Obviously, we would still requite to generate the 3D model of the scene, so basically the 3D model of the car/building/whatever else. Now, instead of having a digital camera positioned at (x,y,z) that takes a 2D image and reshapes it on a 3D form, we have a 3D form what will be “painted” digitally with time-varying elements, and then the outcome will be projected (again digitally) to a 2D surface positioned where the photo camera would be. As such, we obtain a 2D sequence of images, or a video. This is something that can be handled and projected by a real-life projector. And there you have it. Once you have this video, you use a projector, position it exactly where you did for the digital 3D model, and make it run the video. All done. You should now obtain 3D projection mapping.