人体运动捕获

Kinect标定

http://www.ee.oulu.fi/~dherrera/kinect/

但使用Kinect的童鞋也不必沮丧，在OpenNI2.1 beta中，小斤看到了新增的convertDepthToColorCoordinates()方法可以做一些深度与彩色坐标数据的转化，它的效果应该是与device.setImageRegistrationMode( IMAGE_REGISTRATION_DEPTH_TO_COLOR )类似的，有兴趣的童鞋可以尝试一下。

 

The Vitruvian Manifold: Inferring Dense Correspondences for One-Shot Human Pose Estimation

http://research.microsoft.com/apps/pubs/default.aspx?id=162510

Hasler

https://www.mpi-inf.mpg.de/~hasler/

Hannover

http://www.tnt.uni-hannver.de/project/

http://www.ceremade.dauphine.fr/~peyre/teaching/manifold/

Accurate 3D Pose Estimation From a Single Depth Image

http://vis.uky.edu/~gravity/research/mocap/mocap.htm

Pinocchio library for automatic character rigging

http://www.mit.edu/~ibaran/autorig/pinocchio.html

Daniel Vlasic.Articulated Mesh Animation from Multi-view Silhouettes. ACM Transactions on Graphics 27(3), 2008 
http://people.csail.mit.edu/drdaniel/mesh_animation/

Vlasic

http://people.csail.mit.edu/drdaniel/