Hiya folks,
Currently doing some work for a client and they’re wanting to use a 3d model (fbx) with animation baked in to collide with a point cloud generated by our Kinect v2.
Would anyone be able to point me in the right direction of a tutorial or something along the lines of what I’m looking for?
Thanks in advance!
Hi @kaneviggers,
phew - this is a tall order. Firstly wondering what the interaction would be as the pointcloud from the kinect is constantly updating (the points are driven) or is it a saved out state from the kinect?
On the other hand I don’t currently see a path for calculating collisions between an animated geometry and a pointcloud.
Do you have a visual reference for what they are asking?
cheers
Markus
Yeah you make a good point, on reviewing the request, with the points being constantly instantiated creating an interaction between a 3d model and a point cloud being generated by the Kinect doesn’t seem to be possible…
They didn’t supply a reference, they did however provide the 3d model. Attached is a screenshot of the model in blender, the cups rotate on the z axis, the client is looking for an effect where the cups displace the point cloud. Hope that make sense!
Hi @kaneviggers,
still not sure this is doable at all and this might be leading you down the wrong track but you can check out the OP Snippets for the Nvidia Flex Solver, especially the “bowl” example. You could imagine that every element of the provided 3D Model is a Actor and rotating the Actor on the COMP level on it’s z axis.
The Actor that holds the Fluid simulation can contain a Position Feedback TOP which you might be able to use together with the kinect pointcloud. But you also need states in this idea: pre- and post- collision to determine if the pointcloud is driven by the kinect or by the simulation…
Sorry - no idea if this is doable.
cheers
Markus
