Skip to main content

Über dieses Buch

Visualization in scientific computing is getting more and more attention from many people. Especially in relation with the fast increase of com­ puting power, graphic tools are required in many cases for interpreting and presenting the results of various simulations, or for analyzing physical phenomena. The Eurographics Working Group on Visualization in Scientific Com­ puting has therefore organized a first workshop at Electricite de France (Clamart) in cooperation with ONERA (Chatillon). A wide range of pa­ pers were selected in order to cover most of the topics of interest for the members of the group, for this first edition, and 26 of them were presented in two days. Subsequently 18 papers were selected for this volume. 1'he presentations were organized in eight small sessions, in addition to discussions in small subgroups. The first two sessions were dedicated to the specific needs for visualization in computational sciences: the need for graphics support in large computing centres and high performance net­ works, needs of research and education in universities and academic cen­ tres, and the need for effective and efficient ways of integrating numerical computations or experimental data and graphics. Three of those papers are in Part I of this book. The third session discussed the importance and difficulties of using stan­ dards in visualization software, and was related to the fourth session where some reference models and distributed graphics systems were discussed. Part II has five papers from these sessions.



General requirements


1. Scientific Visualization in a Supercomputer Network

Larger amounts of data produced on supercomputers have to be analysed using visualization techniques. As most of the users are not located at supercomputer sites, fast networks are needed to visualize computed results on the desk of the user. Centralized and distributed visualization modes and services, based on video equipment, framebuffers and workstations are discussed. Transfer rates for visualization purposes in local and wide area networks are derived. They are compared to transfer rates between supercomputers and workstations.
U. Lang, H. Aichele, H. Pöhlmann, R. Rühle

2. Visualization Services in Large Scientific Computing Centres

The R&D divisions of EDF and ONERA are two large research centres using supercomputers. They have many similarities in the architectures of their computing facilities and in the way they are operated. In this paper, we summarize common aspects of the management of visualization services in such environments.
Michel Grave, Yvon Le Lous

3. The Visualisation of Numerical Computation

Parallel processing has now made tractable the numerical solution of large complex mathematical models which are derived from across a whole spectrum of disciplines. Such powerful processing capabilities usually result in a vast amount of data being produced. However full value for this advance can only be realised if the engineer has an effective means of visualising the solution obtained. We need to develop efficient and effective ways of integrating numerical computation and graphical techniques.
Lesley Carpenter

Formal Models, Standards and Distributed Graphics


4. Performance Evaluation of Portable Graphics Software and Hardware for Scientific Visualization

In this paper we present an evaluation of the Programmer’s Hierarchical Interactive Graphics System (PHIGS) [2, 3] as portable scientific visualization graphics software, for six different graphics workstations.
Nancy Hitschfeld, Dölf Aemmer, Peter Lamb, Hanspeter Wacht

5. Visualization of Scientific Data for High Energy Physics: Basic Architecture and a Case Study

Visualization of scientific data although a fashionable word in the world of computer graphics, is not a new invention, but it is hundreds years old; examples of Visualization of Scientific Data are dating back since 1700. With the advent of computer graphics the Visualization of Scientific Data has now become a well understood and widely used technology, with hundreds of applications in the most different fields, ranging from media applications to real scientific ones.
In this paper, we discuss the design concepts of the Visualization of Scientific Data systems in particular in the specific field of High Energy Physics, at CERN. Then an example of a practical implementation is given.
“… a computer display enables us to examine the structure of a man-made mathematical world simulated entirely within an electronic mechanism. I think of a computer display as a window on Alice’s Wonderland in which a programmer can depict either objects that obey well-known natural laws or purely imaginary objects that follows laws he has written into his program.
Through computer displays I have landed an airplane on the deck of a moving carrier, observed a nuclear particle hit a potential well, flown in a rocket at nearly the speed of light and watched a computer reveal its innermost workings”.1
Carlo E. Vandoni

6. The IRIDIUM Project: Post-Processing and Distributed Graphics

This paper presents one of the latest project in scientific visualization undertaken by the Direction des Etudes et Recherches, the research and development department of Electricite de France, the French national electricity company. The aims of this project, called Iridium, are to develop a new post-processing tool for visualization in fluid dynamics and more generally to investigate how to take advantage of cooperative architectures including supercomputers and graphics workstations to achieve more powerful systems in the scientific area. The development of the post-processor itself will be the first experiment in which we will be involved in distributing software and we hope that it will be helpful for further applications. We maintain that distributing a scientific application on a supercomputer and one or more workstation connected together through a high speed network ideally responds to the requirements of scientific visualization.
D. Beaucourt, P. Hemmerich

7. Towards a Reference Model for Scientific Visualization Systems

Reference models have been developed in various fields of information processing. The aim of such models is to define a unique basis for system development, system usage and for education and training.
One of the first reference models in Computer Graphics was developed by Guedj et al[9]. Today, a (Standard) Computer Graphics Reference Model is under development by ISO[1]. In areas like CAD reference models have already been established on a national basis[4]. The ermerging Imaging standard[2] also defines a reference model for image operations.
In Scientific Visualization various system models have been presented in recent years. These models focus on different aspects, such as a model for the visualization process, error accumulation, output pipelines in the visualization process, semantics of interaction in the visualization process, architecture and hierarchy of software modules, computing architectures and load sharing models, data and image interfaces.
These models mostly have been set up by users than by developers of tools. Each of the above models does not reflect the meaning of scientific visualization for it’s own but just a certain view. Existing standards, like those known from Computer Graphics (GKS, PHIGS, …) are not covered by these models.
This paper introduces a reference model for visualization systems and classifies existing models using the criteria of this reference model as a basis for comparison.
W. Felger, M. Frühauf, M. Göbel, R. Gnatz, G. R. Hofmann

8. Interactive Scientific Visualisation: A Position Paper

This paper summarises the author’s views on current developments in interactive scientific visualisation. It is based on a talk presented at the Eurographics ’89 conference, held in Hamburg in September 1989. The paper takes issue with the direction of some current work and identifies axeas where new ideas are needed. It has three main sections: data presentation methods, current visualisation system architectures, and a new approach based on parallel processing.
R. J. Hubbold



9. HIGHEND — A Visualisation System for 3D Data with Special Support for Postprocessing of Fluid Dynamics Data

Large and expensive supercomputers are producing an enormous amount of data at significant costs. In order to use these facilities efficiently dedicated peripheral hardware and software is necessary. Computer graphics help the researcher to prepare data for the supercomputer and to process the data produced by numerical solvers.
Hans-Georg Pagendarm

10. Supercomputing Visualization Systems for Scientific Data Analysis and Their Applications to Meteorology

Two supercomputer-based scientific visualization systems implemented over a half-year period are augmented with graphics computers and peripherals. The comparison of these two systems shows that the animation efficiency is dependent upon individual computers: graphics capabilities and computation power; networking and file transfering efficiencies; memory management capacity; and software compatibility. A case is presented where applications of these systems by a meteorologist involves investigation of evolutions of weather systems. Based on his scientific knowledge, the meteorologist selected physically-related parameters for producing three- dimensional animations. The case study results show a composite animation with these related parameters. This animation reveals important weather system development mechanisms which could not have been realized by animations of individual parameters.
Philip C. Chen

Rendering techniques


11. Rendering Lines on Curved Surfaces

The visualization of functions over surfaces is a classical application of scientific visualization. Several techniques, such as grid-lines, contours, and smooth shading, can be employed, each with their own advantages. The result of the combined use of those techniques is often disappointing because the lines, drawn in image space, are unrelated to the surface, rendered in object space. A better result can be achieved if the lines are modelled in three dimensions. A model for lines as semi-transparent tape stuck on the surface is presented. Its integration in rendering systems is discussed, and an implementation and results are presented.
Jarke J. van Wijk

12. Interactive Three-Dimensional Display of Simulated Sedimentary Basins

Simulated sedimentary basins that are represented by volumetric data obtained from a numerical model are visualized by
  • a set of isochron sediment surfaces,
  • a lattice of vertical sections,
  • basement topography, and
  • water body.
Sediment surfaces and sections are color-coded to represent sediment composition in alternative ways, including sediment type and sediment age. It is possible to interactively
  • show or hide any portion of the basin,
  • rotate, resize, and vertically exaggerate the basin,
  • highlight a sediment type,
  • show basin evolution through time, and
  • switch sediment classification methods.
Two viewing programs implemented on top of two different graphics systems (Doré[2] and GL[1]) are described and compared. One viewing application is built from a set of independent but cooperating programs that provide graphic functions and a user interface under X-Windows.
Christoph Ramshorn, Rick Ottolini, Herbert Klein

13. Visualization of 3D Scalar Fields Using Ray Casting

In this paper a high-quality rendering technique for the visualization of 3D scalar fields in an ”aquarium model” is presented. This technique is based on ray casting. The model consists of scalar values defined on a grid, supplemented with glass walls and a bottom. Using trilinear interpolation within the grid is an option. The implementation has been divided into two stages; ray casting, and colour binding using transfer functions are implemented as separate processes. This enables the user to generate different pictures from the same viewpoint, experimenting with different parameters. The method has been applied to data obtained from simulations computing concentrations in sea water.
Andrea J. S. Hin, Edwin Boender, Frits H. Post

14. Volume Rendering and Data Feature Enhancement

This paper describes a visualization model for 3D scalar data fields based on linear transport theory. The concept of “virtual” particles for the extraction of information from data fields is introduced. The role of different types of interaction of the data field with those particles such as absorption, scattering, source and colour shift are discussed and demonstrated.
Special attention is given to possible tools for the enhancement of interesting data features. Random texturing can provide visual insights as to the magnitude and distribution of deviations of related data fields, e.g., originating from analytic models and measurements, or in the noise content of a given data field. Hidden symmetries of a data set can often be identified visually by allowing it to interact with a preselected beam of “physical” particles with the attendant appearance of characteristic structural effects such as channeling.
Wolfgang Krueger

15. Visualization of 3D Empirical Data: The Voxel Processor

Image processing is an area where the application of parallelism is very suitable, because of the large amounts of data involved. In this particular case, 3D images (voxel-images) have to be processed interactively. Operations on the voxel-images involve 3D image processing and visualization from arbitrary angles. The paper describes the development of a prototype voxel-processor, based on a network of T800 transputers.
W. Huiskamp, A. A. J. Langenkamp, P. L. J. van Lieshout

16. Spatial Editing for Interactive Inspection of Voxel Models

Voxel models are suitable for the representation of three dimensional objects of arbitrary topological complexity. They are mostly used for storing spatially sampled real-world data or data resulting from scientific simulation programs. In order to bring out the possibly highly irregular structure of the volume data, a visualization system for voxel-based objects should not only offer various (surface- or volume-) rendering methods, but also spatial editing operations.
We propose using an editing method, based on binary space partitioning. Construction of the binary space partitioning tree, that represents the subdivision of the voxel model, is done by interactive steering of the partitioning planes through the voxel model. The resulting BSP-tree is subsequently used in the rendering of the object. The advantage of a BSP-tree based partitioning is that it may be used in conjunction with many existing volume and surface rendering algorithms.
G. J. Jense, D. P. Huijsmans



17. The Rotating Cube: Interactive Specification of Viewing for Volume Visualization

A part of the user interface of a volume visualization system is described. It provides the opportunity of the real-time interactive definition of viewing parameters for volume rendering. Viewing parameters in this case are the view point and cut planes through the volume data set. It uses an approach for the fast rendering of volume data which traditional computer graphics does not know and which is as fast as wire frame representations.
Martin Frühauf, Kennet Karlsson

18. Chameleon: A Holistic Approach to Visualisation

Scientists from beyond the field of computing are becoming increasingly aware of the advantages to be gained by visualising their problems. Not only does it increase productivity, but if used intelligently it can improve the user’s understanding of the problem.
N. Bowers, K. W. Brodlie


Weitere Informationen