Research On Motion Capture Algorithm Based On 3D Static Model

In recent years,with the rapid development of computer technology,three-dimensional virtual character animation and games have become more and more common in people's daily lives.The 3D virtual character animation in the computer has gradually become the core technology of computer 3D animation technology,virtual reality and augmented reality.At the same time,the real 3D human motion capture technology and the recognition of human motion are in 3D animation and sports competition.Sports analysis and personalized film and television effects production of human characters have a good application prospect.How to construct a prototype system for 3D human motion capture has three important problems:1)3D human body data acquisition technology;2)skeletal skin binding deformation technology;3)human motion recognition technology.Based on this,this paper mainly uses the three-dimensional scanning device Kinect to collect human body model data and bone data,and proposes a new model deformation method and high recognition rate motion motion recognition algorithm,and completes the prototype system design of human motion motion capture.In the model deformation stage,in order to reduce the deformation quality problem in the model motion and improve the efficiency of weight calculation,a model deformation method combining dual quaternion and bounded biharmonic weight is presented.Firstly,the static 3D model is tetrahedralized,then the model's bone control points are set and the weight of each bone segment is calculated.Next,the 3D model is blended with the bone data by the dual quaternion skinning algorithm;Finally,the static 3D model motion is driven by the bone data.Experiments show that this method makes the 3D model have better deformation effect when rotating.In the recognition stage of human motion,the improved dynamic time warping algorithm is used to identify the human motion.Firstly,the Kinect device is used to obtain the skeleton data of the human body motion,and then the distance and angle of the joint point are extracted as the characteristics of the model.Finally,the data collected in real time is matched with the data in the action model library to determine the motion of the human body.