Wrapping digital actor over Depth point cloud to match video
Page 1 of 1

Author:  pashashapiro [ Fri Sep 09, 2016 1:55 am ]
Post subject:  Wrapping digital actor over Depth point cloud to match video

I'm trying to build a system that would use depth sensor capture to reconstruct 3d surface of moving actor accurately enough to be overplayed over RGB camera footage of the same performance.
ipi allows to use to get somewhat accurate skeleton mocap.
However mocap alone is not sufficient to get all the deformations of the surface of the subject.
I'm thinking :
After clean topology model is animated using ipi mocap i could use some sort of wrapping algorithm
to confirm this topology to the point cloud gathered from the depth sensor.
Because mocap animation would bring vertex of clean topology model close enough to corresponding area of the point cloud so that correspondence doesn't need to be manually guided on every frame (perhaps in several places on timeline )

Dose anyone knows of any system that would allow to do something like that?
Does anyone tried to use Ipi to "marry" depth capture with independent(non kinect) RGB camera?

Page 1 of 1 All times are UTC - 8 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group