Virtual Hand Simulation Research With Realistic Motion And Robust Grasping

Virtual hand technique is one of the key technologies to achieve natural and efficient human-computer interaction. And the grasping interaction technology based on virtual hand plays an important role in applications such as human-computer interaction interface and virtual reality, computer animation, etc. The grasping interaction technique is expected to emulate the real situation and therefore provides a "virtual reality" immersion and makes users feel nothing different form the real world. In addition, the grasping interaction techniques should has a realistic motion virtual interactive technology in real time. However, there are contradictions between these two objectives. Therefore, the main part of our work is focus on two aspects of the virtual hand simulation technology, the reality of motion and the stability of grasping. This paper is organized as follows.In the first chapter, Introduction Part, we briefly introduce the status of virtual reality technology and the importance of interaction which based on virtual hand. The related work, such as the virtual hand modeling techniques, bone-skinning algorithms as well as grasping operation, is also introduced. We thus obtain the motivation after analyzing the advantages and disadvantages of the traditional bone-skinning algorithm.In the second chapter, we present a real-time virtual hand deformation algorithm which can conserve the reality of motion. It is the first time that introducing the dual quaternion into the virtual hand deformation algorithm. On the basis of anatomical structure, we construct a corresponding skeleton to achieve the virtual grasping goal.In the third chapter, we model a new hand model for increasing the robustness of finger-based grasping. A model for the realistic and dynamic approximation of the fingers’ stiffness behavior is most important to allow a solid adaption of the contact area. For a robust collision-based soft hand body simulation, we extend the Fast lattice shape matching algorithm, propose a robust manipulation method with applied force for a very precise and robust grasping.The fourth chapter is the focus of this work, in order to increase the robustness of finger-based manipulations and releasing of the virtual objects and perform in real-time with realistic motion, we use free-form deformation for the integration of the virtual hand with realistic motion achieved in second chapter and the robust grasping model based on the applied force achieved in third chapter. Our implementation demonstrates that this approach allows very precise and robust grasping with realistic motion.In the last chapter, we conclude our work, point out the problems that need to be solved and indicate the future work.