Call for Papers
Part VI: Parallel and Distributed Environments
Multi-Resolutional Parallel Isosurface Extraction based on Tetrahedral BisectionThomas Gerstner and Martin Rumpf
Nowadays, multi-resolution visualisation methods become an indispensable ingredient of real time interactive post processing. We will here present an efficient approach for tetrahedral grids recursively generated by bisection, which is based on a more general method for arbitrary nested grids. It especially applies to regular grids, the hexahendra of which are procedurally subdivided into tetrahedra. Besides different types of error indicators, we especially focus on improving the algorithm's performance and reducing the memory requirements. Furthermore, parallelisation combined with an appropriate load balancing on multiprocessor workstations is discussed.
A Volume Rendering Crossbar and SRAM-Based HardwareMiron Deyssenroth, Martijn de Boer, Alexander Gropl, Jurgen Hesser and Reinhard Manner
This chapter introduces an architecture, which combines flexibility and direct hardware support for the shading and compositing part of volume rendering. Flexibility is required for the implementation of different rendering models such as the Heidelberg Raytracer or raycasting. We propose an architecture based upon lookup tables (LUT) and crossbar switches as basic building blocks. This fully pipelined system has been simulated in C and a floorplan has been designed. An operation speed of 40-45 MHz has been estimated. The data path has been limited to 8-9 bits due to the lookup table approach. The error induced by this data path reduction could be limited by increasing accuracy during compositing.
Algorithmically Optimised Real-Time Volume RenderingBernd Vettermann, Jurgen Hesser and Reinhard Manner
One of the most difficult problems in volume rendering hardware architectures is to provide an efficient hardware integration of algorithmic optimisation techniques like, space leaping and early ray termination. An architecture, which solves this problem, is presented in this chapter. It is based on a mutlithreaded approach that overcomes branch hazards caused by these algorithmic optimisation techniques. Additionally, a new memory architecture suitable for the mutlithreaded processing of rays is presented. Simulations show efficiencies of up to 97% (only 3% idle time) for the full system. A frame rate of 132 Hz for a data set of 256x256x128 and up to 4 Hz for 10243 voxels has been found by simulation.
Hardware Accelerated VoxelisationShiaofen Fang and Hongsheng Chen
This chapter presents a hardware accelerated approach to the voxelisation of a wide range of 3D objects, including curves/surfaces, solids, and geometric and volumetric CSG models. The algorithms generate slices of the object models using a surface graphics processor to form the final volume representations. Boolean operations in a volumetric CSG model are carried out using frame buffer blending functions. By storing the resulting volume in the 3D texture memory, the algorithms can also volume rendering the models in real time by 3D texture mapping. As a result, we are able to perform interactive object manipulations and Boolean operations in an intermixed environment of geometric and volumetric objects under a unified volume graphics framework.For images, and relevant information on this and other projects, please visit
Volume Graphics and the InternetKen Brodlie and Jason Wood
The Internet offers opportunities to volume visualisation in two ways: by linking computers, and by linking humans. In this review, we look at distributed processing of visualisation data, particularly via the World Wide Web. We introduce a classification scheme that aims to provide a structure for the field. We also look at collaborative visualisation, where the Internet links humans so that they can co-operate in the analysis of research results. We show that the popular class of Modular Visualisation Environments can be extended to multi-user operation, and we describe in particular an extension of IRIS Explorer which is now available off-the-shelf. For images, and relevant information on this and other projects, please visit
|Page Editor: Andrew S. Winter||Last Updated:1 February 2001|