CARTOGRAPHIC GENERALIZATION IN VIRTUAL REALITY
Alejandro C. FRERY! Cledja K. R. da SILVA? Evandro de B. COSTA* Eliana S. ALMEIDA!
! Departamento de Tecnologia da Informagào — UFAL
Campus A. C. Simóes - BR 104 Norte km 14, 57072-970 Maceió, AL - Brasil
{frery,evandro,eliana} @tci.ufal.br
2Centro de Informática - UFPE
Av. Professor Luis Freire s/n, Cidade Universitária 50740-540 Recife, PE - Brasil
cledja@fapeal.br
KEY WORDS: Cartography, Internet/Web, Application, Generalization, Programming, Simplification, Expert System.
ABSTRACT
This article describes the transposition of cartographic concepts to virtual reality (VR) applications. Versions of carto-
graphic maps can be created by using Generalization. These versions can be implemented for VR applications using
Levels of Detail. Cartographic Generalization is concerned with model visualization that uses twelve operators and
domain-specific knowledge. There is no known transposition of these operators and knowledge to the VR realm. We
present an analysis of these operators in the context of VR. A system for 3D Generalization is proposed. Artificial intel-
ligence techniques are used for both selecting the key objects and applying the operators. This system was implemented
with the JAVA language, modelled with UML and employs a MySQL database.
1 INTRODUCTION
Virtual worlds, virtual reality environments or 3D worlds
can be seen as a computational metaphor of the world where
people and objects can interact. They are mainly used in
entertainment, games and simulators, but they are currently
being used in medicine (body human study), psychology
(treatment of phobias) and arts (virtual museums).
A virtual world can have many objects with varying de-
grees of complexity. The simplest ones are formed by
a single geometric shape, while complex objects can be
formed by organized simple ones. The complexity can be
measured with, for instance, the number of polygons, col-
ors and textures. Because of this complexity, user navi-
gation in worlds with complex objects can be hampered.
Renderization can be slow, causing problems in the im-
mersion experience. Some of the main problems in the
navigation in urban virtual worlds are presented in (Bour-
dakis, 1998), while some solutions are discussed in (Frery
et al., 2002). Among the techniques that can be applied to
solve the navigation problems are based on culling algo-
rithms (Cohen-Or et al., 2000).
Generalization is a abstraction information process. In a
virtual world this process may be responsible for simplify-
ing and/or removing objects considering, for example, the
user position. Generalizations are obtained with Levels of
Detail — LODs, versions of objects in progressive levels of
complexity. The LODs are usually obtained by polygon
simplification; some simplification algorithms can be seen
in (Luebke, 2001). There are many techniques for select-
ing each version or LOD (Constantinescu, 2000); the one
considered here is based on the distance between object
and observer.
This work deals with generalization for virtual worlds built
with VRML (Virtual Reality Modelling Language) (WEB
3D Consortium, 2002, Ramos et al., 1997, Crossley et al.,
1997). The process to generate the generalizations is not
200
exclusive of virtual reality; it is a process used in Cartog-
raphy to produce versions of cartographic maps: Carto-
graphic Generalization.
Section 2 presents the Cartographic Generalization and its
operators; section 3 presents the transposition of concepts
and operators employed in cartographic generalization to
virtual reality; section 4 presents the Generalization Sys-
tem 3D with which we validate our idea; finally, section 5
presents the conclusions and future works.
2 CARTOGRAPHIC GENERALIZATION
Cartographic Generalization may be defined as a set of
proceedings applied for construction and visualization of
models. This generalization aims to improve the interpre-
tation of the information to be showed. To accomplish
this task, the generalization uses operators that will be ex-
plained below.
Cartographic Generalization is employed when new maps
in new scales are needed. It is concerned with the ways the
information is shown (emphasizing, distributing and delet-
ing features). This processing will depend on the cartogra-
pher's knowledge about the requirements and the desired
scale.
Figure 1 presents a map generalized in two ways, the first
considering the topographic features (with emphasis on the
distances and number of objects), while the second consid-
ers the touristic features (enhancing important objects in
the area).
2.1 Operators
To get versions of maps, cartographic generalization uses
twelve operators, applied by the cartographer using domain-
specific knowledge. Each operator is responsible for chang-
ing the way information is presented. These operators are
presented following (Davis and Laender, 1999):
Inter
Figi
OPI
OP2
OP3
OP4
OP6
OP7
OP8
OP9
OP1
OP1
OP1
This
gene:
prese
speci
Syste
