Glyph-Based Multi-field Visualization - Scientific Visualization: Uncertainty, Multifield, Biomedical, and Scalable Visualization - page 122

Glyph-based visualization is one possible approach to realize such a visualiza-
tion of multi-field data (and other chapters of this topic part describe alternative
approaches). A parameterized visualization object is considered—called a glyph (or
sometimes also an icon)—such that certain specifics with respect to its form, e.g., its
shape, color, size/orientation, texture, etc., are given according to data values which
this glyph should represent. A glyph-based visualization is then created by arranging
a certain number of these glyphs across the domain of reference (these could be just
a few, or just one, or many, even so many that they merge into a dense visualization)
such that every glyph becomes a visualization of the data at (or nearby) the location
where the glyph is placed.
Glyph-based visualization approaches span a certain spectrum from, for example,
dense arrangements of relatively simple shapes (stick figures [ 17 ] would be an exam-
ple) to individual instances of complex glyphs that reveal a lot of information (but
only for few, selected places)—the local flow probe [ 13 ] would be an example for this
type of a glyph-based data visualization. Glyph-based visualization approaches also
vary with respect to whether they are constructed in a 2D or 3D visualization space.
We think that it also makes sense to consider glyph-based visualization approaches,
which are based on the placement of glyphs on surfaces within 3D (called 2.5D in
the following). Additionally, we can differentiate visualization solutions according
to which form aspects are varied according to the data, and how many different values
a glyph eventually represents (usually this number is not too large, often 2-4, but
then also examples exist where dozens of values are represented).
A property of all glyph-based visualization approaches is that a discrete visual-
ization is created (instead of a continuous representation like a color map)—only at
certain locations across the domain individual glyphs are instantiated to represent
the data. This means that this approach is only suitable, when it is possible to assume
a certain minimal degree of continuity of the data such that a mental reconstruction
of the data, in particular also in the space between the glyphs, is at least principally
possible. In scientific visualization, this often is possible, making glyph-based visual-
ization particularly interesting for this particular field of application. Alternatively, a
glyph-based visualization also makes sense for discrete data, if a one-to-one relation
between every instance of the data and the glyphs is established.
In the following, we first review a selection of techniques that have been proposed
for glyph-based data visualization. Then, we continue with a discussion of critical
design aspects of glyph-based visualization, not at the least oriented at opportunities
to deal with the perceptual challenge that is inherently associated with this form of
visualization approach.
13.2 State-of-the-Art
This section presents a selection of important papers with a focus on glyph-based
multi-field visualization. A categorization is given based on the visual channels such
as color, shape, size, texture and opacity occupied by the glyph in requirement for