Human Motion Modeling from Complementary Skeleton Joints of Multiple Kinects

摘要

Human motion modeling with the proposed algorithm of complementary skeleton joints can help us to solve the problem of visual blind spot when the motion are captured by Kinects. The shooting model can be remodeled and reconstructed with the recombination and complementation of the skeleton joints by a series of three-dimensional coordinate transformation according to the different Kinects and a virtual monocular camera. The reconstructed human skeleton model can be applied to further motion simulation from the virtual monocular camera.