3D Dynamic Virtual Mirror
The ultimate
objective will be to allow the customer to move their real body and see a 3D image
of their clothed virtual body in real-time .
Input Data
Additionally to data required for the 3D static
solution in the dynamic case we need information to animate the
geometry retrieved from the 3D scanner. We decided to use motion capture data
for this purpose. Since it is not a trivial task to animate the body
automatically we currently use Poser for applying movements to the scans.
An example skeleton animated by applying motion capture data is shown below.
Dynamic Cloth Collision Detection
Like in the 3D Static solution for collision detection and response we
use depth and normal information retrieved from the graphics hardware
after orthogonal rendering. Example depth and normal map videos are
shown below.
Results of the 3D dynamic cloth
model.
|