You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Systems for data processing, anaylsis and representation

Faculty of Computer Science, P.O. Box 4400
University of New Brunswick, Fredericton, N.B. E3B 5A3
email: cware @UNB.ca
Highly interactive 3D graphics is on the threshold of becoming affordable and this opens the door to a whole new range
of applications. However, while a well defined set of interaction techniques exists for the 2D screen environment, the
3D user interface is evolving rapidly. This paper focuses on the user interface to 3D graphics environments. Novel
input devices, 3D widgets, Fly-By techniques, Stereo displays, and Fish Tank virtual reality are discussed within the
framework of J.J. Gibson's theory of affordances.
L'interface utilisateur pour la visualisation en 3D
Le graphisme en 3D hautement interactif est sur le point de devenir abordable et cela ouvre la porte vers toute une
panoplie de nouvelles applications. Néanmoins, alors qu'il existe un ensemble bien défini de techniques pour
l'environnement de l'écran en 2D, l'interface utilisateur en 3Dévolue rapidement. Cet expose va se concentrer sur
l'interface utilisateur pour un environnement graphique en 3D. Des nouveaux périphériques d'entrées, les widgets en 3D,
: z . £ 2.3 . ^ . 2 . A >
les techniques de survol, la représentation en stéreo, la réalite virtuelle vont etre discutés à travers la notion théorique de
J.J. Gibson, 'affordances'.
KEY WORDS: 3D views, Geographical Information System, Visual data display.
highway planning, appreciating sight lines in urban
planning or forestry, and representing ore bodies in
1. INTRODUCTION mining applications are all areas where 3D
visualization can help. In general 3D visualization is
useful where the structures to be understood are
physically displaced in the vertical direction and this
displacement is critical to the application. Three
dimensional representations will become even more
important as GISs evolve into GCADS (Geographical
Computer Aided Design Systems).
The great contribution of Geographical Information
Systems can be seen as an advance from one
dimensional information, usually tables of scalar
coordinates, to two dimensional information - a map
linked to a data base. The methods for manipulating
maps are well established and they consist for the most
part of scale and translate operations for positioning, as
well as point, line and area specification for data input.
A third dimension is added to a limited extent in the
form of overlays, providing layers of information.
However, overlays are a convention for representation,
in general the layer does not denote scalar height
To realize the potential of 3D GIS a number of
problems must be solved. In particular, there are new
data visualization and manipulation techniques to be
invented (after all true 3D output devices are still
many years away). While 2D GIS systems have
borrowed heavily from the paper maps that preceded
them, and which are often the most useful product. 3D
GIS has no such antecedents, unless it is the model of
the landscape architect. There are other problems
related to the need for a the database to support true 3D
visualization and to retain volumetric information, and
we need to support true 3D queries into that database.
These database issues are beyond the scope of this
This may all be about to change as high performance
3D graphics systems become commonplace. There are
increasing numbers of people experimenting with 3D
representations of data and it is becoming clear that this
is useful and practical for many applications. For
example, understanding the surface topography in