3D Video Projection mapping

How are you projecting the image exactly?

Cool video link by the way!

 

I'm going to attempt ASCII art. Here goes...

Code (csharp):

1.  
2.         |  object  |
3.         \          /
4.          \________/
5.          
6.     \
7.      \                /  user
8.       \              /
9.        \            /
10.         \ projector/       
11.  

Ok so seems to me you have a projector and a user. You have a virtual scene (the safe in the other example) which you are pretending is actually occurring inside the object.

So what you would do is exactly reconstruct the scene inside Unity, getting distances etc all perfect. Call this scene 'A'
You would create the object in a temporary material because you are going to texture map over the top of it later.

Next you would create another scene (call it 'B') except you would place the 'fake' scene (the safe stuff) inside the object in this version. You would render this scene from the point of view of 'user' to a texture.

Now you need to get this texture and texture map it back onto the outside of 'object' in scene A. You'd probably end up doing this in a shader, and for every pixel project the 3D point into the 2D camera space of 'user', and look up into the texture you rendered. Then when you look at the scene from the point of view of 'user' in scene 'A', scene 'A' looks exactly the same as scene 'B'.

However render scene 'A' from the point of view of 'projector', and you have a video signal that you can output into your projector, and complete the illusion in real life.


...Hey, there's a lot of words up there, how did I do, did this make any kind of sense?


*edit* just realised that I used the word 'scene' which has connotations in Unity. There might only be one Unity scene, but you would have two 'scenes' within it (I use the word 'scene' to mean 'a collection of stuff' laid out a certain way).

 

Yes: "You would render this scene from the point of view of 'user' to a texture."