【作者】 王亮；

【导师】 戴国忠；

【作者基本信息】 中国科学技术大学 ， 计算机应用技术， 2008， 博士

【摘要】 虚拟现实正在成为一种新的计算手段,它在国防、科学研究、工程设计和制造业等许多重要领域有着广泛的应用前景。从交互的观点看,虚拟现实是一类新出现的三维用户界面(3D UI),与现有的二维WIMP界面有很大不同,面临着许多新的设计挑战和更复杂的可用性问题。现有的3D UI不是过于简易就是缺乏可用性,难以满足实际应用的需求。为此,必须对高可用的三维交互技术进行深入研究;从应用的角度看,缺乏自然、和谐的交互方式已成为制约虚拟现实进一步应用的“瓶颈”,将多通道交互与三维交互相结合,研究多通道三维交互技术已成为研究的热点。论文简要介绍和分析现有的三维交互技术,从设计层次的角度上进行总结;现有的面向通用任务的交互技术还缺乏对应用的支持力度,为此,论文基于场景语义和语义对象的概念,给出场景语义的定义,设计基于场景语义的三维交互技术,使用户通过与语义对象的基本交互实现复杂的应用;并通过具体应用(虚拟家居布局)验证场景语义理论的有效性。采用多通道三维交互是实现自然、和谐的虚拟现实系统的有效途径,因此,论文从多通道输入的角度,给出交互语义的定义,借鉴已有的任务语法分析的理论,提出新的分层整合算法,实现多通道的信息融合;设计基于交互语义的三维交互技术:并通过具体应用(虚拟城市规划)验证交互语义理论的有效性。本文工作中的主要创新可以归纳为以下三点:1.从面向应用的角度,给出场景语义的定义,基于场景语义和语义对象的概念,使用户通过与语义对象的基本交互实现复杂的应用。提出场景语义的设计策略,设计和实现了基于场景语义的三维交互技术。2.从多通道输入的角度,给出交互语义的定义,提出相应的整合算法,实现多通道的信息融合,提出基于时间约束的交互语义整合机制。3.基于场景语义和交互语义设计自然、高效的双手交互技术和多通道交互技术,包括无缝微缩模型技术、动态成组技术、多对象快速选取技术、形状区域对象排列技术。更多还原

【Abstract】 Virtual reality(VR)is becoming a new computing means.It has wide potential applications in many important areas such as national defense,scientific research, engineering design and manufacturing.From the viewpoint of human-computer interaction,VR comes forth as a three-dimensional user interface(3D UI),which is very different from 2D WIMP interfaces and confronts with a lot of new challenges and more complex usability issues.Many existing 3D UIs hardly meet the actual application requirement because they are too simple or lack usability.Therefore,we must carry out more deeply research about high usable 3D interaction techniques (ITs).From the perspective of application,the lack of natural and harmonious interactive modes has become a "bottleneck" restriction for the further application of VR.Based on the combination of multimodal interaction and 3D interaction,the study on multimodal 3D ITs has become a research focus.In this dissertation,the existing 3D ITs are presented,analyzed,and summarized from the viewpoint of design-level.The support provided by the existing common task-oriented ITs to the application is insufficient.Therefore,in this thesis,based on the concepts of scene semantics and semantic objects,the definition of scene semantics is put forward.Some 3D ITs are designed based on scene semantics,which make complex application be carried out by the users’ basic interaction with semantic objects.The validity of scene semantics is proved by a practical application-"virtual furniture layout".Using multimodal 3D interaction is an effective means to achieve natural and harmonious VR systems.Thus,interaction semantics is defined from the view of multi-channel input.A new hierarchical integration algorithm is proposed with reference to the existing theory of task syntax analysis so as to realize multimodal fusion.Some 3D ITs are designed based on interaction semantics,and the validity of interaction semantics is confirmed by a practical application-"virtual city planning".The three innovative ideas in this dissertation are summarized as follows. 1.Scene semantics is defined from the viewpoint of application.Based on the concepts of scene semantics and semantic object,complex application can be carried out by users’ basic interaction with semantic objects.The designing policy of scene semantics is proposed,meanwhile some 3D ITs are designed and carried out based on scene semantics.2.Interaction semantics is defined from the viewpoint of multi-channel input. A corresponding integration algorithm is proposed to realize multimodal fusion.And a mechanism for the integration of interaction semantics is proposed based on time-constraints.3.Based on scene semantics and interaction semantics,natural and effective two-handed 3D ITs and multimodal 3D ITs such as seamless world-in-miniature,dynamic grouping,fast multi-object select and detail-skipped shape arrange are designed.更多还原