Scientific visualization in virtual reality - interaction techniques and application development

Michal Koutek

Data-flow visualization systems working as network editors

Scientific visualization is a well-established method for analysis of data, originating from scientific computations, simulations or measurements. Due to the rapid progress in computer graphics and Virtual Reality (VR), we can see today a number of serious VR applications. Among applications in science also belongs visualization of data in virtual environments (VEs). The potential of VR for three-dimensional visualization seems obvious. However, practically it is still very difficult to interact with virtual environments. Development of VR applications and design of VEs are also issues that need more attention. Visualization and exploration of data in VEs requires development of new visualization concepts different from those known from desktop workstations. These problems became motivation for this research project and this thesis. The practical VR implementation has been performed on the Responsive Workbench (RWB). During the project we developed the RWB Library and RWB Simulator, our application software development environment. On top of this we designed and implemented VRX, an interactive visualization toolkit for the Workbench. Chapter 2 contains a survey of related work in the field of visualization in VR, describing general principles of scientific visualization, the state of the art in Virtual Reality, concentrating on the research issues of scientific data visualization by means of VR. This chapter also gives a motivation of our work and outlines the research agenda of this thesis. Chapter 3 introduces the Virtual (or Responsive) Workbench, also briefly describing its technical aspects. Further, the design aspects of applications and VEs for the Workbench are discussed. Interesting aspects are the laboratory table metaphor and the window-on-world metaphor. The technical constraints and Workbench specifics are reflected in the RWB Library. The concepts and usage of this library are described in detail. Implications of the development of RWB applications with the RWB Simulator are also discussed in the chapter. Chapter 4 deals with interaction techniques in virtual environments. After related work in this field, interaction techniques specifically suitable for the Workbench are presented. This includes selection and manipulation of virtual objects, and navigation and exploration of VEs. The VR aspects of object collisions and constraints during object manipulation are also discussed. The problem of providing force feedback on the Virtual Workbench has been addressed by a force feedback method for manipulation of virtual objects that can be used as an alternative to haptics. The spring-based manipulation tools (spring, spring-fork and spring-probe) are introduced. These tools are based on the spring metaphor, providing visual force feedback. The spring tools have been further adapted into the spring manipulator, which has been used for particle steering (manipulation with atomic particles) in a real-time Molecular Dynamics simulation, visualized on the Workbench using the MolDRIVE system. Three methods for particle steering are described and demonstrated with example applications. Chapter 5 presents our approach to exploration and data visualization in VR. The description begins with our interaction and exploration tools. One- and two-handed interaction scenarios are discussed. The tools, derived from the scenarios, were based on several real-world metaphors, which made these tools very intuitive and easy to use. The navigation and probing tools developed were incorporated in the VRX toolkit, of which the concepts and techniques are also described in detail. VRX is our modular object-oriented toolkit for exploratory data visualization, and it enables rapid development of visualization applications for the Workbench. A multiprocessing scheme and adaptive resolution of the probing tools is used for enhancement of performance. Chapter 6 presents several case studies, dealing with visualization of scientific data on the Workbench. Three main case studies are described: interactive visualization of flooding scenarios, Molecular Dynamics visualization and computational steering, and visualization of cumulus clouds. The cumulus clouds case study served a test-bed application during the development of VRX. For the Molecular Dynamics case study we developed MolDRIVE, a system for visualization and steering of MD simulations. This system has been used for studies of the β-Alumina electrolyte and several proteins. This thesis shows which aspects are necessary for the development of virtual environments for visualization of scientific data. Interaction with the VEs is studied extensively. It is clearly demonstrated by object manipulation and by particle steering that visual force feedback, based on the spring metaphor, is a good alternative approach to the physical force feedback (haptics).


More Information

Citation

Michal Koutek, Scientific visualization in virtual reality - interaction techniques and application development, PhD Thesis, Delft University of Technology, 2003.

BibTex

@phdthesis{bib:koutek:2003,
    author       = { Koutek, Michal },    
    title        = { Scientific visualization in virtual reality - interaction techniques                   and application development },
    year         = { 2003 },
    school       = { Delft University of Technology },
    dblp         = { books/daglib/0008706 },
    url          = { https://publications.graphics.tudelft.nl/papers/558 },
}