A Motion Retargeting Research Based On Inverse Kinematic

In recent years,with the development and popularization of motion capture technology,more and more motion capture data have emerged,resulting in a large number of complete mo-tion capture databases,providing valuable research materials for the computer animation field,making the role of action.The study has entered a new period.The motion editing technol-ogy of character animation has attracted the interests of domestic and foreign researchers,and in recent years,a relatively mature technology framework for motion redirection has gradually emerged.However,in general,research in this field has not been particularly perfect.The ex-isting algorithms are mainly for the study of free-role movements,and there are few studies on space and its own constraints.The algorithms are universal and specific to some interaction with the environment.The study of the motion distortion that appears is less.This article mainly summarizes the various inverse kinematics and motion redirection methods that are already in existence.The related research on the interaction between motion and terrain appears to have the following work:First introduces and analyses the inverse kinematics and the comparison of the existing main movement redirection algorithm,and discusses the adaptability between different algorithm and the advantages and disadvantages,a variety of different methods for the inverse kinematics algorithm.For 3D human body skeleton model,this paper proposes a motion retargeting method about steps constraints.Through the constraint of joint end of the chain,we can solve the problem of exercise redirects the steps in the process of sliding,and kept footsteps on the ground position.At the same time,we made some steps for position in time and space constraints,and damping least square method of inverse kinematics for added control item,to get more reasonable results.In this article,we use the real motion capture data in the experiment.The motion data of human body character is analyzed,and the real motion capture data in real production is used.In order to solve the problem of distortion of the ground in the process of human motion,the inverse kinematics is solved,and the bone model with different proportion of joints has better effect.In the case of 3D terrain,an adaptive algorithm is proposed.At the same time,in order to smooth the motion of the joint,reduce the beating of each joint position,add the smoothing of the data,so that the motion between different frames is more continuous.Experimental results show that this method can effectively solve the human role in the process of the movement to redirect to the ground of all kinds of distortion,at the same time to ensure the processing results to keep the original motion data in main movement characteristic,makes the result more similar to the movement of the capture data.