Welcome to VG99
Call For Papers
Style Guide (Paper)
INVITED INDUSTRIAL SPEECHES
OpenGl Volumizer: Breaking Down the Barriers in Volume VisualizationROBERT GRZESZCZUK
Silicon Graphics, Inc., USA
Despite sustained demand for volume graphics in a number of rapidly growing markets, substantial benefits from its use, and active research in the field, volumetric objects do not enjoy the prevalence their surface counterparts do. Arguably, this is due to monolithic and inflexible designs that plague many existing systems, unorthodox approaches that evolve in isolation from mainstream graphics solutions hindering convergence, the lack of a viable unified framework for treatment of both surfaces and volumes, and finally, disregard for market forces which necessarily impact adoption of any new technology.
The presentation will focus on a framework for representing volumetric, or solid, objects in a way that is consistent with the representational scheme widely accepted for surface-based objects. In the process, we decouple geometry from appearance for volumetric objects, introduce the concept of Volumetric Primitives, and propose a volume graphics pipeline that lends itself to powerful yet economical software and hardware implementations. Finally, we describe our implementation of the concepts we introduce and illustrate how easily this flexible scheme can be cast onto ubiquitous hardware using OpenGL Volumizer as an example.
Robert Grzeszczuk is a graphics engineer at Silicon Graphics and is currently involved in architecting OpenGL Volumizer API--a high level open toolkit addressing visualization issues common in the geoscience and medical markets. His professional interests include computer graphics, visualization, image processing, and computer vision in which areas he authored, or co-authored, more than 60 abstracts and articles. Prior to joining SGI he was on faculty of the Department of Radiology at the University of Chicago. He holds a BS in Mathematics, MS, and PhD in EECS from the Electronic Visualization Laboratory at the University of Illinois at Chicago.
The VolumePro Real-Time Ray-Casting SystemHANSPETER PFISTER
MERL -- A Mitsubishi Electric Research Laboratory, USA
In this talk we present VolumePro, the world's first single-chip real-time volume rendering system for PC class computers. Based on the Cube-4 architecture developed at the State University of New York at Stony Brook, this system fits on a PCI board and is capable of rendering a volume data set of 16 mega-voxels (e.g., 256^3 datasets) at 30 frames per second.
VolumePro implements object-space ray-casting with parallel slice-by-slice processing similar to shear-warp rendering. Our discussion of the architecture focuses on the rendering pipeline, and the memory organization. We describe several advanced features of VolumePro, such as gradient magnitude modulation of opacity and illumination, supersampling, supervolumes, and crop and cut planes. The system renders more than 500 million interpolated, Phong illuminated, composited samples per second. We will report on the chip implementation, and give an overview of our plans for commercial development of real-time volume graphics.
Hanspeter Pfister is a Research Scientist at MERL - A Mitsubishi Electric Research Laboratory in Cambridge, MA. He is the chief architect of VolumePro, Mitsubishi Electric's real-time volume rendering system for PC-class computers. His research interests include computer graphics, scientific visualization, computer architecture, and VLSI design. Hanspeter Pfister received his PhD in Computer Science in 1996 from the State University of New York at Stony Brook. In his doctoral research with Prof. Arie Kaufman he developed Cube-4, a scalable architecture for real-time volume rendering. He received his Dipl.-Ing. degree in electrical engineering from the Department of Electrical Engineering at the Swiss Federal Institute of Technology (ETH) Zurich in 1991. He is a member of the ACM, IEEE, the IEEE Computer Society, and the Eurographics Association.