ANIMATION
THEORY AND ITS APPLICATION TO SPATIO-TEMPORAL ANALYSIS (Menno-Jan
Kraak)
Connie
Blok - Dynamic visualization variables in animations; Their use for
the monitoring of spatio-temporal data
ITC
Department of Geoinformatics, Cartography Division
P.O.Box 6 7500 AA Enschede, The Netherlands
e-mail: Blok@ITC.NL
For
various reasons, earth scientists are interested in the monitoring
of spatial dynamics. Spatial dynamics result from changes in the characteristics
of spatial objects in time. One way to capture spatio-temporal data
is by periodic earth observations, and change information can be detected
from those data through application of statistical and/or visual methods.
An advantage of the use of visual methods is that objects can be viewed
in their spatial context. And if computer animations are produced,
e.g. from (processed) remotely sensed time series data, they can also
be viewed in their temporal context. In animated maps or images the
spatial dimensions of a static map or image can be used with the graphic
variables. In addition, there is a temporal dimension (display time),
in which a number of dynamic variables can be applied. Several dynamic
variables have been distinguished so far by DiBiase et.al. (1992)
and MacEachren (1994), and they are also applied. However, some questions
remain to be answered, such as: are these the basic types, what are
their characteristics and relationships, what is the relation to the
graphic variables, how can they be successfully applied, and for what?
Some of these will be addressed in the proposed research.
The
research will focus on the application of dynamic variables in 2D
computer animations. The perspective is that of a domain specialist
who is involved in the processing of remotely sensed data for monitoring
purposes in a GIS-environment. The specialist wants to reveal unknowns,
and the visualization goal is exploration. Although animations are
very suitable for the display of dynamics, exploration of changes
is not easy. Many changes may occur within display time, and all over
the area. So it is likely that a viewer will miss some of them. Not
seeing is a common visualization error (MacEachren & Ganter, 1990).
In the research the author wants to investigate whether not seeing
the changes in time series can be reduced if the dynamic variables
are used as real design instruments in computer animations. In a first
phase, the dynamic variables and their characteristics will be defined.
Then animations will be produced in which the dynamic variables are
manipulated in several ways, and their ability to assist domain specialists
in the detection of spatial dynamics will be tested. Expected results
of this phase are a theoretical framework for application of the dynamic
variables for the monitoring of spatial dynamics in remotely sensed
time series, and recommendations for the development of animation
tools that allow relevant manipulations of the dynamic variables by
domain specialists in a GIS environment. Prerequisites for a possible
second phase are that the tools are actually developed and a link
between the visualization environment and the GIS data base is provided.
In the second phase of the research then, it will be tested whether
and how the available tools are actually used by the domain specialists.
The result of this phase will be further recommendations for improvement
of the exploration environment.
References:
DiBiase, D., A.M. MacEachren, J.B. Krygier & C. Reeves (1992),
Animation and the role of map design in scientific visualization.
In: Cartography and Geographic Information Systems, Vol. 19, No. 4,
pp. 201-214, 265-266
MacEachren,
A. M. (1994), Visualization in modern cartography: setting the agenda.
In: A.M. MacEachren & D.R. Fraser Taylor (eds), Visualization
in modern cartography. Modern Cartography, Volume two. Oxford: Elsevier
Science Ltd., pp. 1-12 MacEachren, A.M. & J.H. Ganter (1990),
A pattern identification approach to cartographic visualization. In:
Cartographica, Vol 27. No. 2, pp. 64-81
Gerd
Buzieck - Experiences in the design of legends for cartographic animations
http://visart.ifk.uni-hannover.de/~buziek/COMVIS/ifkvis.html
Alan
MacEachren & Donna J.Peuquet - Integrating GIS and Visualization
for Analysis of Space/Time Environmental Data.
This
presentation provides on overview of the Apoala Project, an effort
to develop and assess a prototype Temporal Geographic Information
System with integrated multivariate spatiotemporal visualization capabilities.
Emphasis is on one aspect of the project dealing with the concept
of "operations" defined at multiple levels of the system.
The
Apoala prototype emphasizes analysis of dynamic environmental processes.
It includes a multi-representational database model which combines
object-based, location-based and time-based representations of environmental
data in a tightly coupled manner linked to a multidimensional visualization
capability to facilitate exploratory and task-based spatio-temporal
analysis. Three premises underlie our approach to this prototype:
attention to time as well as space is essential to understanding environmental
(and other) processes; treating time as an attribute of space,or as
a 4th dimension does not capture all relevant aspects of time; and
design of systems that facilitate full understanding of environmental
processes requires consideration of time, location, and features from
an integrated perspective.
Most
past approaches to GIS have been largely atemporal -- at best treating
time as an attribute equivalent to other coverages/layers in the database.
Change has been addressed by comparing time snapshots. Cartographic
approaches to representing the world have also typically been atemporal,
with the small-multiple style "chess" map being equivalent
to temporal snapshots in a GIS. As with GIS, change has often been
addressed by comparing time snapshots, with the result depicted in
a "change" map. Process has, occasionally, been addressed
cartographically through "dance" maps that represent change
in object location and map animation used to represent change in either
locations or attributes over time -- but most animations use the "snapshot"
conception of time.
Our
contention is that a complete perspective on representing geographic
phenomena must address their dynamic nature in a fundamental way --
thus treating space, time, and objects as co-equal, but different,
aspects of geographic information. The figure below presents an approach
for doing just that. further details on this integrated approach to
modeling space-time information can be found in: Peuquet, D.J. (1994).
It's about time: A conceptual framework for the representation of
temporal dynamics in geographic information systems. Annals of
the Association of American Geographers, (in press): 441-461;
and Peuquet, D.J. and Duan, N. (1995). An event-based spatiotemporal
data model (ESTDM) for temporal analysis of geographical data. International
Journal of Geographical Information Systems, 9(1): 7-24.
For
the prototype, we are addressing system design at three levels [following
an approach outlined in: Howard, D. and MacEachren, A.M. 1996. Interface
design for geographic visualization: Tools for representing reliability.
Cartography and Geographic Information Systems, 23(2): 59-77].
The levels are: conceptual ‹ at which what and who the system
is for are considered; operational ‹ at which conceptual goals
are sub divided into a set of discrete operations applicable to the
data; and implementational ‹ at which methods for achieving the
operational goals are addressed (within particular, hardware, software,
and problem context constraints). Within the operational level, our
initial step is to develop a taxonomy of operations that takes into
account (a) the user's cognitive sub-division of broader conceptual
goals into discrete operations, (b) the visual display as an interface
for initiating operations and seeing their result, and (c) the formalization
of operations necessary for linking user actions with the systems
database.
This research is supported under a grant to the authors from the U.S.
Environmental Protection Agency [#: R825195-01-0]
VISUALIZATION
OF UNCERTAINTY (Alan MacEachren)
Lucy
Bastin, Pete Fisher and Jo Wood - Visualization of fuzzy spatial information
in spatial decision making
Department
of Geography, University of Leicester
Leicester LE1 7RH United Kingdom
phone: +44 (0) 116 252 3839 fax: +44 (0) 116 252 3854
email: lb26@le.ac.uk, pff1@le.ac.uk, jwo@le.ac.uk
Much
has been written amongst spatial theorists on the use of fuzzy sets
in the storage, interpretation and analysis of spatial information.
There would appear, however, to have been few attempts to develop
approaches to conveying this type of uncertainty to the users of spatial
information, and no reported results which show its effectiveness.
We are embarking on a project with EU funding on the extraction, verification
and visualization of fuzzy land cover information from satellite imagery.
There are two major components to the visualization. First is the
use of advanced systems to support the process of information extraction,
and especially the linking of imagery, attribute and classification
in multiple multi-dimensional spaces, and its exploration. This module
is expected to employ object linking within a virtual reality environment,
and is being developed at the Joint Research Centre in Italy. Of immediate
concern to the present authors is the second stage of visualization,
the interpretation of the extracted fuzzy information and its delivery
to the user community.
The
problem can be stated as conveying to a user of the land cover information
the fact that the land cover classification which may be derived from
a satellite image is a matter of interpretation, and that the interpretation
is dependent on the user and the application. Two alternative approaches
to this problem may be taken. In the first, the user would be presented
with the original satellite reflectance data, and the complete classification
system, enabling extraction of the optimal classification for the
application. However, application scientists and decision makers are
more familiar with information classes, and so it is believed that
the user should be presented with a database relating to the fuzziness
of information classes together with a visualization environment for
the viewing and the analysis of those classes, so that the scientist
can develop their own specialist interpretation.
Many
sources for confusion exist among theorists and users of spatial information
in the use of this information. First, as mentioned above, is the
education of the information users that the classic Boolean map is
only one possible interpretation. Much support and education will
be required for this conceptual shift in the user community. Second,
a number of fuzzy logic operators will need to be available to the
user to support their interpretation since it may well involve the
combination of elemental fuzzy information classes. This will require
a flexible variety of operators, with appropriate user assistance
and worked examples. Finally, a variety of methods of visualization
of partial membership imagery have been proposed, including random
animation, serial animation, grey scale imagery, and pseudo 3-D projections.
A range of these should be available to the user, but they will need
to be presented with an initial, interesting image to foster their
use and exploration of the possibilities inherent in the dataset.
The decision as to which view should be presented will dictate the
acceptance of the approach amongst users, and so is perhaps the most
crucial design decision.
This
work is in its earliest stages and the system architecture has yet
to be decided. It is to be hoped that attendance at this meeting will
help focus the methods and objectives. It is hoped that at least two
of the researchers on this project will be able to attend the meeting,
although no paper was submitted or has been accepted at the main ICA
meeting.
GEOREFERENCED
DESKTOP VIRTUAL REALITY (David Fairbairn and Antonio Camara, the latter
helped organize activities on this theme, but was not able to attend
the meeting)
Theresa-Marie
Rhyne- Future Reality: Visualizing Integrated Scientific Information
ACM
SIGGRAPH
Director at Large Lockheed Martin Technical Services
US EPA Scientific Visualization Center
86 Alexander Drive Research Triangle Park, North Carolina 27711
Tel: (919) 541 - 0207 Fax: (919) 541 - 0056 Email: rhyne@siggraph.org
Scientific
visualization (SciVis) transforms numerical or symbolic data and information
into geometric computer generated images. It is a methodology for
interpreting image data entered into a computer as well as data generated
from computational models. In general, SciVis is based on the application
of techniques from the convergent fields of: computer graphics; image
processing; computer vision; computer-aided design, signal processing
and user interface design studies. For geographic data, this includes
assisting with spatial analysis and integrating with geographic information
systems (GIS). The rapid development of the World Wide Web (Web) facilitates
collaborative sharing of visualizations among remotely located users.
Tools like the Virtual Reality Modeling Language (VRML) support three
dimensional web visualizations on heterogenous computer platforms.
Traditionally,
SciVis tasks have been associated with high performance computing
activities. Recent advances in hardware, computer graphics libraries
standardization, and the networking infrastructure begin to allow
for real time visualization on Pentium personal computers. It is now
possible to imagine a future where multi-dimensional geographic data
can be accessed via the Web and visualized in real time with a personal
computer. Is this our future reality?
Kate
Moore (presented by Jason Dykes) - Using Java to Interact with Geo-Referenced
VRML
Department
of Geography, University of Leicester
Leicester LE17RH, UK
Email: mek@le.ac.uk, jad7@le.ac.uk, jwo@le.ac.uk
Virtual
reality technology is providing earth scientists and cartographers
with new, exciting and interactive ways to model the world and real-life
phenomena. One of the most important functions of traditional cartography
is providing information about location: where the user is, where
an object is located and what is at a location. This is equally, if
not more, important for navigating virtual worlds and referencing
information from the real world.
The Virtual Field Course (VFC) is developing software to enable students
to experience fieldwork environments through multimedia, including
virtual reality. The VFC will exploit enhancement, interaction, extension
and replacement potential of virtual reality within the context of
the aims of residential fieldwork.
As a preliminary to developing a spatial interface for a Virtual Field
Course geo-referenced multimedia database, a 2-dimensional Java interface
has been developed to provide information and interaction with 3-dimensional
VRML terrain models. During this process both the scope and the constraints
of geo-referenced VRML have been examined.
Bin
Jiang - Realistic Visualisations of Urban Environments
Bin
Jiang, Martin Dodge, Andy Smith, Simon Doyle and Mike Batty
Centre for Advanced Spatial Analysis University College London
1-19 Torrington Place, London WC1E 6BT, UK
Tel: + 44 171 391 1255, Fax: + 44 171 813 2843 Email: b.jiang@ucl.ac.uk
The
Virtual ENvironments for Urban Environment (VENUE) project, funded
by the Joint Information Systems Committee(JISC), aims to provide
computer based tools to facilitate urban design and planning process.
Desktop GIS is used as a basic platform which links some other plug-in
models such as space syntax analytical tools and virtual reality(VR).
This presentation began with a brief introduction of VENUE project,
followed by the abstract visualisation of urban environments with
which the traditional cartography is concerned. The presentation focused
on the realistic visualisation, particularly with dynamic and interactive
3D VRML models and photorealistic rendering by demonstrating a number
of examples in the context of urban design and planning. Finally,
the presentation is concluded by pointing out that the combination
of both abstract and realistic visualisation is a better choice for
the purpose of analysis and design, that the integration of GIS and
VR technology would greatly facilitate urban design process, and that
the integration of WWW and GIS would help conduct collaborative planning
via internet.
Jorge
Nelson Neves- Virtual Environments and Geographical Representation
Environmental
Systems Analysis Group
Department of Environmental Sciences and Engineering
New University of Lisbon, PORTUGAL
e-mail: jnn@mail.fct.unl.pt
The
use of Virtual Environments or Virtual Reality techniques hasthe potential
for spatio-temporal representations with great fidelity,because the
users have the possibility to interact directly and immersively with
the elements which form the basis of their decision. This abilityto
feel the results of an action, as opposed to qualify them, makes the
understanding clearer and perhaps universal.
Virtual
reality technologies are available now at reduced cost and its useis
becoming common in the World Wide Web (WWW). With the advent of theVirtual
Reality Markup Language (VRML) and multi-user environments, desktop
VR became attractiveparticularly in collaborative environments. Thus,
not only the technicaland scientific communities will benefit from
its application to spatialproblems; citizens anywhere in the world
will have a chance to participate interactively in the solutions for
such problems.
Real
time visualization of large Digital Terrain Models (DTM),along with
proper interaction techniques, are crucial to understandcomplex, constantly
changing conditions and to prototype experiments using simulation
models.
The
data used to represent a DTM usually has giant proportions(from several
megabytes to terabytes). No system is capable of processingthat information
in real time. Considering that we can not visualize allthe information
at once, the computer should only process the data relevant to theimage
it generates each frame. Thus, one should have several models withlevels
of complexity depending on the observation distance, a concept termed
Level-Of-Detail (LOD) modeling.
In this
research project the different terrain and texture representations
for each scale aredetermined using a mathematical tool to hierarchically
decompose functions- Wavelets. The discrete data at several resolutions
is stored in aquad-tree and managed together with the LODquad-tree
to allow multiscale simulation and visualization i.e., theability
to update and visualize data at several resolutions simultaneously.
Dave
Fairbairn - Generalization of VRML data
Dept.
of Geomatics, University of Newcastle upon Tyne
Newcastle upon Tyne NE1 7RU, England
Tel: +44 (0)191 222 6353. Fax: +44 (0)191 222 8691 e-mail: Dave.Fairbairn@ncl.ac.uk
Large
file sizes for VR data sets encoded in VRML lead to difficulties in
download times over the Web and in handling data in browsers. Use
of the concept of Levels of Detail in VRML is an important mechanism
to assist in rendering detailed data at varying `scales' (i.e. distances
from the viewpoint), although its use does in fact make file sizes
larger as the different representations are held together in one VRML
file. Automated methods of creating the different levels is sought.
The latest version of one browser (Pioneer Pro) claims that it has
such an automated tool and public-access software also exists: e.g.
Lodestar (from Technical University of Vienna, Computer Graphics research
group). Only valid for early VRML version 1.0 data sets, Lodestar
can read in IndexedFaceSet or IndexedLineSet (e.g. polygon coverages
or wire frame diagrams). Successive levels of embedded IndexedFaceSets
are attached to LOD nodes in the VRML file whilst different properties
of faces and lines (e.g. texture, materials, normals etc.) are maintained.
The algorithm used is based on octrees (note that some cartographic
generalisation uses the 2D equivalent, quadtrees) with adaptive depth.
Successive decompositions of the 3D structure into cubes is undertaken
and nodes within each cube are collapsed to one, which then has pointers
to the child nodes valid for the different Levels of Detail attached
to it. The collapsing is undertaken using a number of heuristic algorithms
which calculate factors for each node. The 3 algorithms used in Lodestar
address differing measures in assessing the generalization efficiency.
Examples were shown of both the standard Lodestar model (a galleon
ship) and a Vistapro-created terrain model at successive generalisation
levels.
MULTIMEDIA
& VISUALIZATION (William Cartwright)
Alexandra
Koussoulakou - The use of Multimedia in Department of Cadastre, Photogrammetry
and Cartography of the University of Thessaloniki
Department
of Cadastre, Photogrammetry and Cartography
The Aristotle University of Thessaloniki Thessaloniki - Greece kusulaku@eng.auth.gr
At the
Department of Cadastre, Photogrammetry and Cartography of the University
of Thessaloniki there is a growing interest regarding the use of mulitmedia.
In the Cartography and Photogrammetry groups, in particular, some
projects have been and are being carried out. These have to do with
the production of software for supporting the educational process
(tutorials in Photogrammetry and Cartography), for the documentation
of cultural monuments and the presentation of subjects related to
the history of cartography.
Also
within the whole Faculty (Rural and Surveying Engineering) there is
interest for setting up a project for the production of multimedia
educational software to be used in the educational process, which
will probably cause more interest for multimedia applications in the
near future. The author is involved with the cartographic part of
the above (and partly with the photogrammetric aspects, since she
also gives a course in Introductory Photogrammetry). It began a couple
of years ago with setting up a cartographic tutorial, concerning the
subject of cartographic animation. At the moment this work (i.e. the
tutorial) is being revised and extended in order to cover another
topic of cartographic interest, namely the use of mulitmedia in Cartography,
by means of hypermaps. Both topics within the tutorial (animation
and hypermaps) are using examples of various types of maps in order
to demonstrate theory. For the hypermap part a subject of historical
interest was chosen (dispersion of population through time). The tutorial
is currently under development.
Michael
P. Peterson - Maps and Multimedia on the World Wide Web
Department
of Geography &Geology University of Nebraska at Omaha
geolib@cwis.unomaha.edu
The
number of maps distributed through the World Wide Web seems to have
surpassed print and CD-ROM. More important than numbers, is the form
in which the maps are distributed and the level of interaction that
is presented to the map user. The web seems to be defining, more than
any other media, the form of map delivery and its multimedia content.
The web interface is shaping the map-user interaction and has important
implications for how maps will be presented in the future. Three stages
of interaction in web-mapping are examined: static, interactive, and
animated. File size, affecting the speed of access, seems to be the
major criteria in the development of each form of map presentation.
The major trend seems to be in developing new forms of interaction.
A smaller emphasis is placed on content. Examples are presented through
Netscape.
William
Cartwright - Multimedia Extensions to Geographical Information Depiction
Research and Development at RMIT
Department
of Land Information RMIT University Melbourne VIC Australia w.cartwright@rmit.edu.au
The
paper provides an overview of the impact of electronic publishing
and communications in undergraduate courses in Cartography, Geomatics
and Surveying at RMIT. It also gives descriptions of research and
development in multimedia extensions to geographical information being
undertaken by graduate students and academic staff members.
Projects
discussed are:
-Design
rules for multimedia map products
-Virtual Atlas on the World Wide Web
-The use of the WWW to distribute GIS-based tourist information
-Development of a Virtual Worlds interface
-GeoExploratorium
Future
plans and proposed research directions are outlined.
