Issue No. 04 - July-August (1998 vol. 18)
Visualization is very much an application-driven field. Without applications to motivate, use, and evaluate visualization methods, they would not have much value except possibly for their visual aesthetics. Today, we see an ever increasing and diverse application of visualization ranging from more traditional areas such as medicine, weather, and aeronautics to more recent and exotic areas such as bioinformatics and Web visualization.
The rich variety of application areas produces a healthy mix and vital influx of ideas that contribute to and improve the current set of visualization techniques. For example, isosurfaces and direct volume rendering have their roots in medicine, while streamlines originated from aeronautics. The beauty of this synergy is that apart from helping advance the application area with more powerful analysis tools, these tools have enriched the tools in our trade. Plus they have been extended to applications other than their original intended use.
In this fashion, visualization grows by absorbing, adapting, and applying ideas and techniques from a variety of fields including computer graphics (shading and texturing), image processing and vision (gradient detection and smoothing operations), and mathematics (quaternions and splines). Experiments and field testing by practitioners in the field as reported in case studies further improve the usefulness of these methods and serve as guides for others.
In This Issue
Visualization often paves the way for rapid advances in an application. Consequently, progress in technology for visualization speeds up advances in the application area. Progress in scientific visualization may affect the methods used to map invisible data or structures to visible objects on the screen. It may also help us interact with the visualization of the underlying data or simulation intuitively.
Each year, the IEEE Visualization conference devotes part of the program to papers illustrating the impact of visualization on a field of application. A selection of the best contributions from these case studies clearly illustrates the statements above. Examples include the visualization of optical phenomena tightly coupled with the simulation of optics, and advanced interaction concepts by linking visualization to aerodynamic databases and integrating them into an augmented virtual reality system. The selection also covers highly specialized visualization techniques such as vortex core detection, visualization of car crash simulations, and improving the perception and effectiveness of advanced rendering methods applied to flow fields.
Some visualization techniques serve the specialized needs of scientists in our universities while other techniques have to prove their robustness in large-scale research projects or industrial design and production environments. The emerging field of interactive visualization presents a great challenge in educational systems or multimedia instructional documents. It is here that visualization meets—much like books do—the untrained, unskilled consumer. Visualization in the context of education should have an even higher awareness of the dangers of misinterpretation and errors.
For a while, visualization provided the glitz to convey the fascination of science and technology and help justify the money spent to achieve it. At the same time, visualization has become one of the essential tools helping scientists, engineers, and specialists access and understand data. Today and in the future we see visualization systems and methods at the hand of everybody in everyday life.
Visualization can provide comfortable access to a world in which abstract structures, networks, data, and information play an essential role in daily business, private life, education, and recreation. We see visualization systems becoming widely available—for example, through visualization services on the Web, which allow anybody to map their data into images. This will make it easier to reuse information and raw data. It will open up opportunities to present data in several ways with different concepts in mind, comparing data from various sources and thereby producing representations and context with a new quality that represents a value on its own.
Visualization is not really ready for prime time in the sense that anybody off the street (or the Web) can use it, fully understand the presentation, and analyze their data correctly. For this to happen, several key steps and components must be developed further. Applications will continue to drive the development of new visualization techniques. The list below focuses on issues that may be overlooked or are of secondary concern to the applications, but nevertheless prove important for visualization to stand up under widespread general use.
- Visualization that facilitates comparisons and verification. To a certain extent, some degree of inaccuracy exists in most visualizations today. That is, our current visualizations tend to gloss over inaccuracies in the data—those introduced in the data derivation process—as well as those from the visualization process itself. To be credible, our visualizations must at least provide users with indications of these inaccuracies. Better yet, they should be able to validate their data in the face of these inaccuracies and measure and compare differences across data sets.
- Evaluation of visualization techniques. Novice users find it difficult to decide which visualization technique suits a particular data set or visualization task. Even veteran users may not exactly know the effectiveness of different visualization techniques. We need to have a systematic evaluation of different techniques for different classes of visualization tasks and data sets. The result could be in the form of a "rule-of-thumb" or more formal data-task-technique mapping.
- Perceptual issues in visualization. Not only are we not exploiting the full potential of presenting information to the user, we sometimes do this incorrectly—as with the well-known example of improperly using linear color maps. This is perhaps the most challenging issue in this list, since it involves cross-disciplinary cooperation. It is also one where visualization is not the solution provider, but rather the client.
The articles in this issue grew out of representative papers from the IEEE Visualization 97 Case Studies. We hope you enjoy them.
Alex Pang is an associate professor in the Computer Science Department at the University of California, Santa Cruz. He obtained his MS and PhD in computer science from UCLA in 1984 and 1990, respectively. His current research interests are in collaboration software, uncertainty visualization, scientific visualization, and virtual reality interfaces. He co-chaired the IEEE Visualization Case Studies in 1996 and 1997, and the SPIE Visual Data Exploration and Analysis conference in 1998. For more information, visit www.cse.ucsc.edu/~pang.
Hans-Georg Pagendarm is head of the Software Technology Group in the Institute of Fluid Mechanics of the German Aerospace Research Establishment (DLR) in Göttingen, Germany. His research interests cover visualization as well as fluid dynamics. He received his engineering degree and his PhD as a mechanical engineer at the Ruhr University at Bochum in Germany. He also holds a diploma from the von Karman Institute for fluid dynamics in Brussels, Belgium. He co-chaired the case studies of the IEEE Visualization conference in 1995, 1996, and 1997. He is chairman of the Eurographics working group on visualization.