节点文献

基于灵巧手的虚拟抓取技术研究

Research on Virtual Grasp Based on Dexterous Hand

【作者】 陈飞飞

【导师】 彭群生; 万华根;

【作者基本信息】 浙江大学 , 计算机应用技术, 2009, 硕士

【摘要】 随着虚拟现实的发展,自然和谐的人机交互日益成为一个重要的研究领域。在基于虚拟现实的训练、教育、娱乐、建筑设计等领域,人机交互有着非常重要的作用。但是,由于计算机生成的虚拟环境无法精确地模拟真实物理世界中的许多规则和约束,因此实现自然、和谐的人机交互非常困难。虚拟手交互技术是人机交互研究领域的研究热点。虚拟抓取是用户与虚拟环境中的物体进行交互的一种自然直观的方式。本文围绕如何在虚拟抓取过程中增强用户体验和减少用户负担展开研究。本文通过对人手解剖结构以及运动学特征的分析,首先提出了一个包含皮肤层、运动层、碰撞检测层、力觉层的4层灵巧虚拟手模型。该模型在保证交互实时性的前提下,提高了虚拟交互的真实感。其次,在对现实世界中的物体进行形状抽象的基础上,提出了基于几何条件的抓取准则,既提高了抓取判别的速度,又减少了误抓现象的发生。此外,在抓取过程中利用虚拟抓取有限状态机进行控制,避免了虚拟手手指嵌入物体。最后,通过对人手抓取轨迹规律的分析,提出了一种自动抓取方法,可以有效地减少用户交互的负担。实验结果表明了本文算法的有效性。

【Abstract】 With the development of virtual reality, natural and harmonious human-computer interaction has become an important field of study. HCI plays a very important role in virtual reality-based simulation, training, education, entertainment, architectural design and other fields. However, the virtual environment generated by the computer can not accurately simulate the real physical world in many of the rules and constraints, it is very difficult to achieve a natural and harmonious human-computer interaction.Virtual hand is one of the hottest topics in human-computer interaction. Virtual grasp is a natural intuitive way for the user to interact with the virtual objects in virtual environment. We present methods on how to enhance the user experience and reduce the burden on the user.We propose a 4-layer flexible virtual hand model, which consist of skin layer, kinematics layer, collision detection layer and haptic layer, for virtual hand interaction. The model can not only enhance the realism of the virtual interaction, but also ensure the premise of real-time. Based on the shape abstraction of objects, geometry based grasping rules are proposed. It can reduce the incidence of false grasp, as well as increase the speed of grasp identifying. We can control the grasping states by using virtual grasping finite state machine to prevent the interpenetration of the virtual hand and objects. In order to alleviate the users’ burden, an automatic grasping method is proposed. The experiment results show the validity of our algorithm.

  • 【网络出版投稿人】 浙江大学
  • 【网络出版年期】2011年 S2期
节点文献中: