The Automated Synthesis Of The Interactive Role Of Action And Implementation

Human motion has played a very important role in file production, video games, virtual reality applications, and control of humanoid robot. It's very difficult to produce high quality human motion animation or design related algorithms. Therefore, in recent years, methods of motion capture based become a hot spot in character animation studies.However, this character animation based on data-driven usually do not have the interactivity with virtual environments. Even through the establishment of motion capture databases,with a relevant selection algorithm can achieve a certain degree of interaction, but this interaction is also considerable limitations. It's not only heavily dependent on the size of the database, but also tequired quite large amount of time when calculation, and thus can not achieve a satisfactory real-time. The advancement of interactive and real-time performance will undoubtedly has considerable significance to the character animatipn producers.The real-time synthesis of the character motion animation with realism and interaction is the main research contents of this issue. We have put forward a complete set of character animation sunthsis methods based on dynamic model, dynamic simulation and dynamics-based motion control, balance control, and so to build a dynamic character animation systhesis syetem, Anima, as the application of research findings and the promotional platform.To enhance the research starting point of the researchers for dynamic character animation, as well the capability of the animation producers to ues the dynamics to create interactive character animation, we presents a dynamic character animation synthesis system, Anima, with the design ideas of core-plugIn. And accordingly to implement a dynamic character animation systhesis editor, Anima Studio, can used as research and development platform, as well as for the systhesis of complex, multi-character scene animation.In the thesis, we describes the process of character dynamics-based modeling through skeleton models, motion representation and collision models. Proportional derivative control method is adopt to simulate the character's muscular force, and increasing the stability of character motion through the use of critical damping and gyroscopic effects compensation. In the motion simulation, we design the force handing and collision processing module, combined with the character's dynamic model, muscle simulation and sensor, ultimatedly form a complete set of dynamics-based character motion simulation method.Character motion control based on dynamics, because of its interaction with the environment, is no longer able to fully comply with pre-set motion strategy. And the trajectory of the character motion will change with environment changing, or outside interaction. Therefor, it requires balance control maintain mechanism laying on the character motion control to maintain the balance when character standing, locomotion and disturbed. In this thesis, finite state machine is designed with balance feedback algorithm to achieve the character balance control mechanism, and using feedback error learning to further improve the stability of the character response action.Finally, through the integrated use of data-driven animation and various motion controllers realized base on dynamic motion control technology, as well as the transition between them, and using non-linear way to edit the animation resources, so as to achieve interactive automated character motion synthesis.