sitemap

contact

University of Stuttgart

	Institute for Visualization and Interactive Systems

Isosurface Extraction Techniques for Web-based Volume Visualization

The reconstruction of isosurfaces from scalar volume data has positioned itself as a fundamental visualization technique in many different applications. But the dramatically increasing size of volumetric data sets often prohibits the handling of these models on affordable low-end single processor architectures. Distributed client-server systems integrating high-bandwidth transmission channels and Web-based visualization tools are one alternative to attack this particular problem, but therefore new approaches to reduce the load of numerical processing and the number of generated primitives are required. In this paper we outline different scenarios for distributed isosurface reconstruction from large-scale volumetric data sets. We demonstrate how to directly generate stripped surface representations and we introduce adaptive and hierarchical concepts to minimize the number of vertices that have to be reconstructed, transmitted and rendered. Furthermore, we propose a novel computation scheme, which allows the user to flexibly exploit locally available resources. The proposed algorithms have been merged together in order to build a platform-independent Web-based application. Extensive use of VRML and Java OpenGL-bindings allows for the exploration of large-scale volume data quite efficiently.

Advanced isosurface representation

1. Reconstructing stripped surfaces

One way to compress a triangle mesh is to reorganize multiple triangles into one triangle strip. For each vertex presented after the first two vertices one triangle is defined. Obviously, organizing triangle meshes as strip sets considerably reduces the number of vertices needed to represent the mesh and thus the number of operations necessary to render the mesh. By providing a method that is capable to directly reconstruct stripped isosurfaces we simultaneously address the problem to reduce the load on the transmission channel and to optimize the rendering performance.

Images

Stripped isosurfaces: cell stripped, single direction stripped, multi-direction stripped

multi-direction stripped isosurface of an MRI scan

Movies(MPEG)

Transferring and visualizing strips at different levels of detail (3.9MB) - OpenGL bindings visualization

2. Level-of-detail isosurface reconstruction

Although in the previous approach the number of primitives was reduced in a quite efficient way by reorganizing the geometric representation, usually interactivity cannot be achieved particularly for large-scale data sets. In general, we believe that in the proposed environment interactive frame rates can only be achieved by taking advantage of the hierarchical nature of multi-resolution techniques to effectively decimate the number of primitives needed to represent the surface.
Adaptive on-the-fly reconstruction and rendering is performed from a hierarchical octree representation thus allowing the user to interactively browse through all possible isosurfaces.
In order to allow for the exploration of even the highest resolution data sets the user can manually select a region of interest in which the surface is reconstructed at the finest level. Outside of this region the surface is reconstructed at increasingly coarser resolution. In this way it is possible to adaptively select the size and the number of details that should be reconstructed at the same time achieving sufficient frame rates for interactive navigation.

Images

focus point (red ball) with small and big radius

focus point (red ball)showing left part of the isosurface with increased level-of-detail

Movies

Moving the focus point on the surface of a head (6.9MB) - VRML visualization

Distributed isosurface reconstruction

We can exploit locally available computing power of the client system to reduce the idle times and to balance the load on all involved units more equally. Therefore we developed a distributed computation scheme which allows vertex positions to be calculated on the client.

Images

CT scan dataset (512x512x120): view onto the eyeballs (aneurysm marked by focus) , aneurysm shown with high level-of-detail

Movies

CT dataset (512x512x106): Moving the focus point on the cerebral artery to an aneurysm (berry-like blister of the cerebral artery) (6.9MB, 13.5MB)

References

R. Westermann, L. Kobbelt, T. Ertl
Real-time Exploration of Regular Volume Data by Adaptive Reconstruction of Iso-Surfaces
The Visual Computer 1998