but also the novice looking for general back-
ground information on visualizat ion topics.
I. Introduction
Part I looks at basic algorithms for scientific
visualization. In practice, a typical algorithm
may be thought of as a transformation from
one data form into another. These operations
may also change the dimensionality of the data.
For example, generating a streamline from a
specified starting point in an input 3D dataset
produces a 1D curve. The input may be repre-
sented as a finite element mesh, while the output
may be a represented as a polyline. Such oper-
ations are typical of scientific visualization
systems that repeat edly transform data into dif-
ferent forms and ultimately into a representa-
tion that can be rendered by the computer
system.
II. Scalar Field Visualization: Isosurfaces
The analysis of scalar data benefits from the
extraction of lines (2D) or surfaces (3D) of con-
stant value. As described in Part I, marching
cubes is the most widely used method for the
extraction of isosurfaces. In this section,
methods for the acceleration of isosurface ex-
traction are presented by the various contribu-
tors. Yarden Livnat introduces the span space, a
representation of acceleration of isosurfaces.
Based on this concept, methods that use the
span space are described. Han-Wei Shen pre-
sents a method for exploiting temporal locality
for isosurface extraction, in recognition of the
fact that temporal information is becoming in-
creasingly crucia l to comprehension of time-de-
pendent scalar fields. Roberto Scopigno, Paolo
Cignoni, Claudio Montani, and Enrico Puppo
present a method for optimally using the span
space for isosurface extraction based on the
interval tree. Koji Koyamada and Takayuki
Itoh describe a method for isosurface extraction
based on the extrema graph. To conclude this
section, Ross Whitaker presents an ov erview
of level-sets and their relation to isosurface
extraction.
III. Scalar Field Visualization: Volume
Rendering
Direction scalar field visualization is accom-
plished with volume rendering, which produces
an image directly from the data without an
intermediate geometrical representation. Arie
Kaufman and Klaus Mueller provide an excel-
lent survey of volume rendering algorithms.
Roger Crawfis, Daqing Xue, and Caixia Zhang
provide a more detailed look at the splatting
method for volume rendering. Joe Kniss,
Gordon Kindlmann, and Chuck Hansen de-
scribe how to exploit multidimensional transfer
functions for extracting the material boundaries
of objects. Martin Kraus and Thomas Ertl de-
scribe a method by which volume rendering can
be accelerated through the precomputation of
the volume integral. Finally, Hanspeter Pfister
provides an overview of another approach to
the acceleration of volume rendering, the use
of hardware methods.
IV. Vector Field Visualization
Flow visualization is an important topic in sci-
entific visualization and has been the subject of
active research for many years. Typically, data
originates from numerical simulations, such as
those of computational fluid dynamics (CFD),
and must be analyzed by means of visualization
to provide an understanding of the flow. Daniel
Weiskopf and Gordon Erlebacher present an
overview of such methods, including a specific
technique for the rapid visualization of flow
data that exploits hardware available on most
graphics cards. Gordon Erlebacher, Bruno
Jobard, and Daniel Weiskopf de scribe their
method for flow textures in the next chapter.
Ming Jiang, Raghu Machiraju, and David
Thompson then provide an overview and solu-
tion to the problem of gaining insight into flow
fields through the localization of vortices.
Preface xv