Game Development Reference
In-Depth Information
Consequently, modelling the road network becomes geometrical as well as topological
(orientation in a network) and semantic (the meaning of the red light) (Thomas, 1999).
Though we have quickly shown that modelling a virtual environment is not purely
geometric, it is nonetheless an important step of the process of constructing a virtual
will have a look at some theoretical models related to the geometric description and
then we will see the modelling tools. The non-geometric aspects will be covered later
in part II “Models for sensorimotor rendering'' and part III “Models for behavioural
rendering''.
14.1.1 Types of objects
Before modelling geometric objects, it is necessary to answer one question: what are we
trying to model? It is obvious that the model of a non-deformable solid is different than
that of a compressible fluid. The simplest object is of course the non-deformable rigid
solid. This object can only undergo solid transformations T
(R, t), where R is rotation
in space and t is translation. The cube in figure 14.2 is an example of non-deformable
rigid solid. The main advantage of rigid solids is that all we need to calculate is their
solid position in the space before rendering them. The non-rigid models, on the other
hand, can be divided into various categories:
=
Articulated solids: These are non-deformable rigid solids connected to each other
by mechanical connections (for example a pivot). They are used to represent for
example, of course as an initial approximation, the movement of an arm connected
to a trunk or the wheels of a car that are connected to the chassis. The pendulum in
figure 14.3 is a similar example. The connections between the solids are represented
by their parameters (here, angles θ 1 and θ 2 );
The deformable solids whose deformations follow the laws of mechanics (or do
not follow them, in case of an imaginary world). For example, the skin of a living
being or a ball that changes its shape when it touches the ground. These models
are controlled by their mechanical parameters. It is thus necessary to calculate the
changes in their shape before carrying out their visual rendering. The images in
figure 14.4 show the changes in the shape of the Luxo lamp under the effect of its
own weight (Cozot, 1996);
Lastly, the non-solid models which represent everything that cannot be associated
with the notion of an object. These include fluids, the special models that connect
multiple sub-models connected mechanically to each other. For example, special
models are used for creating a lava fountain.
There are also some other models which do not come within the scope of virtual reality
for professional use; they are related to the techniques called non-photorealistic ren-
dering (Gooch & Gooch, 2001). These techniques often produce spectacular results,
but they cannot be classified into simple groups like the groups we just mentioned.
Besides, their use in virtual reality is negligible these days. Hence, we will not discuss
them in this topic.
The major constraint of virtual reality in terms of graphic rendering is obviously
the absolute necessity of creating a real time rendering of the 3D model. It is thus