Geoscience Reference
In-Depth Information
3 A Model of 3D Information Visualization Techniques
3.1 Knowledge Sources
We studied the scientific literature about applications based on 3D city models to obtain
a global view of the domain. Although the techniques used in these applications are
often not explicitly described, they provided enough information to draw initial classi-
fication axis. The studied models and applications were used for various tasks, such as:
• Evaluation of the wind comfort for pedestrians in a city street (Amorim et al.
2012 ) where 3D coloured polylines (colour representing wind velocity) are
added to the geometrical model.
• Assessment of air quality in a street or neighbourhood by adding coloured solid
objects to the 3D buildings (Lu et al. 2009 ; San José et al. 2012 ).
• Estimation of vehicle trafic from Yatskiv and Savrasovs ( 2012 ) by adding 3D
objects (vehicles) to “animate” the movement paths.
• Analysis of pedestrian behaviour (Marina et al. 2012 ) where colored bars visu-
alize spatial distribution of pedestrian movement.
• Analysis of human perception of space (Fisher-Gewirtzman 2012 ) where
colored lines represent visual exposure or visual openness in the 3D city model.
• Visualization of historically enriched 3D city models where information (text
and images) has been added to the geometrical model (Alamouri and Pecchioli
2010 ; Hervy et al. 2012 ).
• Emergency evacuation of buildings where a routing network is superposed to
the building (Atila et al. 2013 ).
This study showed that visualization techniques could be classified along three
main axes: the kind of data to be visualized (the input data); the visual rendering
(how data are displayed in the 3D scene); and the usage of the technique, in terms
of context and task.
3.2 Data Representation
The data types are various, ranging from rich text, such as in a pdf file where
images can also be present (Hervy et al. 2012 ), to 3D scalar values (temperature) or
to 3D vector values (wind) (San José et al. 2012 ). An important point to note is that
these data are spatialized. Each data element is associated to some spatial region.
In a vector field each vector is associated to a 3D point, in an energy consumption
dataset each value (in kWh) is associated to a building, in the output of a computa-
tional fluid dynamics model each value is the average fluid speed in a 3D cell, etc.
Moreover, all the coverage regions of an input dataset may be located within a
specific spatial object, as is the case for a vector field on a surface, or measurement
values at points forming a grid.
Search WWH ::




Custom Search