Inbetweening For Human Motion Control And Blending In Computer Animation

Computer animation is an offspring of computer technique and animation production art. With the developing of computer technique, computer animation becomes to one important subject of advanced and new technique area. Motion control method in computer animation mainly includes key-frame, motion capture and the way based on physics. Methods of motion blending include motion signal processing, the way based on statistic model and motion graph. In order to reflect motion true to nature, motion velocity is considered in each way. In this thesis, computer animation inbetweening is proposed combine with motion capture, motion edit and so on. Velocity, essential part of motion, becomes to one quantum which could be control in the production of animation. It improves third dimension in kinetic animation.This thesis presents inbetweening method in parameter curves and implicit curves based on inbetweening. Given parameter curves and implicit curves equation of human end effectors'trajectory, use inbetweening to calculate end-effectors'position in each frame cycle. These frames in the position compose to frame sequence, and could represent figure's movement really. Human motion control based on inbetweening is proposed in this thesis. Selected joint angle as parameter of motion control, and velocity is blend in calculation of each manipulative parameter via inbetweening. The serial frames reflect motion in different velocity excellently. In the human ambulant control, base angle is presented to make motion control more expediently. Inbetweening is introduced to motion blending. Through projective translation, motion of space is decomposed to motion of planes in which we use inbetweening method to deal with motion pose data synchronously. Then, motion of planes is blended to motion of space. This method avoids the problems bring by complicate matrix translation. In this thesis, motion composition based on inbetweening is proposed. We decompose the motion into several sub-motions which have there own rules separately and use synchronized inbetweening to create target motion via them. This method could create composite motion directly based on motional rules which is known and avoid synchronization in motion composition.Several simulations are demonstrated in thesis by MATLAB, and the result proves the feasibility of these methods.