Motion Synthesis Based On Motion Capture Role And Realization

Manual editing role motion based key-frame requires deep understanding of real human motion features, and the editing process is too cumbersome. The reality of Movement data generated from Physical role motion synthesis is very low. Along with the development of motion capture techenologies, role motion synthesis based on motion capture becomes a research directioin of computer role animation field. Also, because of the high cost of motion capture process, the editing of motion capture data has become a new focus.First we introduce the research and development of role animation production, analysis's the background of motion synthesis based on motion capture data, which includes virtual human skeleton model, motion capture data format and inverse kinematics theory. Then, we deeply study and solute the problems of motion data edition in three aspects. They are:(1)Deeply analyses the connection problem of motion serial, and uses linear interpolation and quaternion interpolation to dealing with the problem. Solved human body position change problem with linear interpolation, and pos smoothly change problem with quaternion interpolation. Motion data smoothly connecting problem will be handled with overlay interpolation or transition interpolation.(2)The joint constraint problem at the end of motion capture data is solved by a simplified inverse kinematics method. With fixed ankle joint position, the position and rotation quaternion of knee joint and leg joint has been successfully handled. And the key frame data of the constrained end joint has been solved. Finally, key frame data achieves a smooth transition.(3)Solved motion editing problem based spacetime constraints in the low-dimensional Degrees of Freedom(DOF) space. We get a reasonable mathematical quantification of the spacetime constraints, and establish a reasonable mathematical model of objective function. Optimization method of DOF has been successfully reduced the space dimension of constained conditions and objective function. It greatly reduced the complexity of solving the optimization. At last, through a constraint processing with a normal walking motion data, generated a bow body walking movement through the obstacle.Finally, based on the role motion Synthesis theory, with C++,OpenGL,FLTK and QT, we developed and realized Anima role animation Synthesis system. This system realized a variety ways of synthesis of the role of movement.