3D Human Skeleton Synergy Kinematics Based on the Video

In this paper, based on the automatic extraction of the three-dimensional human skeleton, the kinematics for joint points of 3D human skeleton is introduced. The adjacent joint points show all kinds of motions by the change of position. The conformal geometric algebra method is used to describe 3D human model to realize the direct calculation using geometric language. The human skeleton gesture is driven by the joint points to get the key frames and interpolation between frames to realize gesture sequences. Then, analysis the synergy rules of adjacent joint points. Finally, to getting the better match between 3D human skeleton and the corresponding video. Meanwhile, the joint point´s 3D information makes a good groundwork for recognition the different gestures.