Once the vision system has
determined where the human is, and the agent simulations have figured out
where the creatures should be, it's time to render the scene and show
it to the user.
The depth information determined by the vision system can be combined with the
bitmap of the user and the graphics output of the agent simulations to produce occusions. If the
user passes in front of a creature the creature is occluded, if the
user is behind the creature the user is occluded:
Once the occlusions are determined, the video signal from the camera can be blended in hardware with the graphics output to obtain the desired result.
Graphics shares CPU time with vision system and the agent simulations