State of the Art in Computer Graphics Aspects of Visualization This is the fourth volume derived from a State of . . . the Art in Computer Graphics Summer Institute. It represents a snapshot of a number of topics in computer graphics, topics which include visualization of scientific data; modeling; some aspects of visualization in virtual reality; and hardware architectures for visu alization. Many papers first present a background introduction to the topic, followed by discussion of current work in the topic. The volume is thus equally suitable for nonspecialists in a particular area, and for the more experienced researcher in the field. It also enables general readers to obtain an acquaintance with a particular topic area sufficient to apply that knowledge in the context of solving current problems. The volume is organized into four chapters - Visualization of Data, Modeling, Virtual Reality Techniques, and Hardware Architectures for Visualization. In the first chapter, Val Watson and Pamela Walatka address the visual aspects of fluid dynamic computations. They discuss algorithms for function-mapped surfaces and cutting planes, isosurfaces, particle traces, and topology extractions. They point out that current visualization systems are limited by low information transfer bandwidth, poor response to viewing and model accuracy modification requests, mismatches between model rendering and human cognitive capabilities, and ineffective interactive tools. However, Watson and Walatka indicate that proposed systems will correct most of these problems.

Today one of the hardest parts of computer aided design or analysis is first modeling the design, then recording and verifying it. For example, a typical vehicle such as a tank, automobile, ship or aircraft might be composed of tens of thousands of individual parts. Many of these parts are composed of cylinders, flats, and simple conic curves and surfaces such as are amenable to modeling using a constructive solid geometry (CSG) approach. However, especially with the increasing use of composite materials, many parts are designed using sculp tured surfaces. A marriage of these two techniques in now critical to continued development of computer aided design and analysis. Further, the graphical user interfaces used in most modeling systems are at best barely adequate to the required task. Critical work on these interfaces is required to continue pushing back the frontiers. Similarly, once the design is modeled, how are the varied and diverse pieces stored, retrieved, and modified? How are physical interferences prevented or eliminated? Although considerable progress has been made, there are still more questions and frustrations than answers. One of the fundamental problems of the 1990s is and will continue to be modeling. The second problem is interpretation. With the ever increasing computational power available, our ability to generate data far exceeds our ability to interpret, understand, and utilize that data.

In the third paper in this chapter, Mike Pratt provides an historical intro duction to solid modeling. He presents the development of the three most freqently used techniques: cellular subdivision, constructive solid modeling and boundary representation. Although each of these techniques devel oped more or less independently, today the designer's needs dictate that a successful system allows access to all of these methods. For example, sculptured surfaces are generally represented using a boundary represen tation. However, the design of a complex vehicle generally dictates that a sculptured surface representation is most efficient for the 'skin' while constructive solid geometry representation is most efficent for the inter nal mechanism. Pratt also discusses the emerging concept of design by 'feature line'. Finally, he addresses the very important problem of data exchange between solid modeling systems and the progress that is being made towards developing an international standard. With the advent of reasonably low cost scientific workstations with rea sonable to outstanding graphics capabilities, scientists and engineers are increasingly turning to computer analysis for answers to fundamental ques tions and to computer graphics for present~tion of those answers. Although the current crop of workstations exhibit quite impressive computational ca pability, they are still not capable of solving many problems in a reasonable time frame, e. g. , executing computational fluid dynamics and finite element codes or generating complex ray traced or radiosity based images. In the sixth chapter Mike Muuss of the U. S.

State of the Art in Computer Graphics Aspects of Visualization This is the fourth volume derived from a State of . . . the Art in Computer Graphics Summer Institute. It represents a snapshot of a number of topics in computer graphics, topics which include visualization of scientific data; modeling; some aspects of visualization in virtual reality; and hardware architectures for visu alization. Many papers first present a background introduction to the topic, followed by discussion of current work in the topic. The volume is thus equally suitable for nonspecialists in a particular area, and for the more experienced researcher in the field. It also enables general readers to obtain an acquaintance with a particular topic area sufficient to apply that knowledge in the context of solving current problems. The volume is organized into four chapters - Visualization of Data, Modeling, Virtual Reality Techniques, and Hardware Architectures for Visualization. In the first chapter, Val Watson and Pamela Walatka address the visual aspects of fluid dynamic computations. They discuss algorithms for function-mapped surfaces and cutting planes, isosurfaces, particle traces, and topology extractions. They point out that current visualization systems are limited by low information transfer bandwidth, poor response to viewing and model accuracy modification requests, mismatches between model rendering and human cognitive capabilities, and ineffective interactive tools. However, Watson and Walatka indicate that proposed systems will correct most of these problems.

In the third paper in this chapter, Mike Pratt provides an historical intro duction to solid modeling. He presents the development of the three most freqently used techniques: cellular subdivision, constructive solid modeling and boundary representation. Although each of these techniques devel oped more or less independently, today the designer's needs dictate that a successful system allows access to all of these methods. For example, sculptured surfaces are generally represented using a boundary represen tation. However, the design of a complex vehicle generally dictates that a sculptured surface representation is most efficient for the 'skin' while constructive solid geometry representation is most efficent for the inter nal mechanism. Pratt also discusses the emerging concept of design by 'feature line'. Finally, he addresses the very important problem of data exchange between solid modeling systems and the progress that is being made towards developing an international standard. With the advent of reasonably low cost scientific workstations with rea sonable to outstanding graphics capabilities, scientists and engineers are increasingly turning to computer analysis for answers to fundamental ques tions and to computer graphics for present~tion of those answers. Although the current crop of workstations exhibit quite impressive computational ca pability, they are still not capable of solving many problems in a reasonable time frame, e. g. , executing computational fluid dynamics and finite element codes or generating complex ray traced or radiosity based images. In the sixth chapter Mike Muuss of the U. S.

Today one of the hardest parts of computer aided design or analysis is first modeling the design, then recording and verifying it. For example, a typical vehicle such as a tank, automobile, ship or aircraft might be composed of tens of thousands of individual parts. Many of these parts are composed of cylinders, flats, and simple conic curves and surfaces such as are amenable to modeling using a constructive solid geometry (CSG) approach. However, especially with the increasing use of composite materials, many parts are designed using sculp tured surfaces. A marriage of these two techniques in now critical to continued development of computer aided design and analysis. Further, the graphical user interfaces used in most modeling systems are at best barely adequate to the required task. Critical work on these interfaces is required to continue pushing back the frontiers. Similarly, once the design is modeled, how are the varied and diverse pieces stored, retrieved, and modified? How are physical interferences prevented or eliminated? Although considerable progress has been made, there are still more questions and frustrations than answers. One of the fundamental problems of the 1990s is and will continue to be modeling. The second problem is interpretation. With the ever increasing computational power available, our ability to generate data far exceeds our ability to interpret, understand, and utilize that data.

Reviews 8 key areas which are focal points for current de- velopments. Theseare Design, Modeling, Image Generatiion, Workstations, VLSI, HCI, Graphics Standards, and Electronic Documentation.