Visualization and interaction metaphors for large high-resolution displays [Elektronische Ressource] / von Sebastian Thelen

Sujets

• Informatik

Informations

• Publié par: technische_universitat_kaiserslautern
• Publié le: 01 janvier 2011
• Nombre de lectures: 52
• Langue: English
• Poids de l'ouvrage: 23 Mo

Extrait

Visualization and Interaction Metaphors
for Large High-Resolution Displays
Vom Fachbereich Informatik der Technischen Universitat Kaiserslautern
zur Erlangung des akademischen Grades
Doktor der Ingenieurwissenschaften (Dr.-Ing.)
genehmigte Dissertation
von
Dipl.-Inf. Sebastian Thelen
Datum der wissenschaftlichen Aussprache: 9. Februar 2011
Dekan: Prof. Dr. Arnd Poetzsch-He ter
Vorsitzender des Prufungsausschusses: Prof. Dr. Peter Liggesmeyer
1. Berichterstatter: Prof. Dr. Achim Ebert
2. Berich Prof. Dr. Jorg Meyer
3. Berichterstatter: Prof. Dr. Hans Hagenc Copyright 2010
by Sebastian ThelenAcknowledgements
This thesis would not have been possible without the help of many people who
supported me during my doctoral studies. First of all, I would like to thank my
advisors Achim Ebert and J org Meyer. Achim and J org gave me the opportunity
to freely conduct my research and work on the projects I was most interested in.
In particular, I would like to thank J org for making my visit to the University of
California, Irvine an unforgettable time in both scienti c and non-scienti c matters.
I am especially thankful to Prof. Hans Hagen, who gave me the opportunity to
conduct my research as a member of the International Research Training Group
"Visualization of Large and Unstructured Data Sets - Applications in Geospatial
Planning, Modeling, and Engineering" (IRTG 1131), University of Kaiserslautern,
Germany. More than anything else before in my still short academic life, I enjoyed
the opportunity to work in an interdisciplinary and international environment.
Furthermore, I would like to thank all my friends and colleagues at the University
of Kaiserslautern and the University of California, Irvine. My thanks go to Peter-
Scott Olech for the excellent collaboration which resulted in major parts of this
thesis. I would also like to thank Aaron Knoll for our common work on direct
volume rendering optimizations. Without him, major parts of this thesis would not
exist. Further thanks go to Jan Malburg for his help on the Tiled++ prototype and
the AnyScreen framework. My thanks go to members of the California Institute
for Telecommunication and Information Technology (Calit2); in particular, I would
like to thank Stephen Jenks, Sung-Jin Kim, Falko Kuster, and Kai-Uwe D orr for
allowing me to use parts of their HIPerWall software. I want to thank my o ce mate
Li-Chang Cheng for the fruitful discussions and nice lunch breaks. Furthermore, I
would like to thank Ariane Middel for proofreading my papers and for providing
me with the Maricopa County census data set. Special thanks go to Inga Scheler
and Mady Gruys for helping me manage all the confusing and sometimes annoying
organizational and administrative things that need to be done.
Finally, and most importantly, I would like to thank my family for their con dence
and constant encouragement during the long years of my education. I deeply appre-
ciate my mother’s unwavering support which allowed me to undertake my studies inthe very rst place. I thank my father for having an open ear and for encouraging
me to carry on when times were hard. My thanks go to my grandfather for keeping
me with both feet on the ground. I want to thank Annette for her love, con dence
and telling me to follow my heart. She reminded me that there is a life outside
academia and helped me to not turn into a complete nerd.Abstract
Due to remarkable technological advances in the last three decades the capacity
of computer systems has improved tremendously. Considering Moore’s law, the
number of transistors on integrated circuits has doubled approximately every two
years and the trend is continuing. Likewise, developments in storage density, network
bandwidth, and compute capacity show similar patterns. As a consequence, the
amount of data that can be processed by today’s systems has increased by orders
of magnitude. At the same time, however, the resolution of screens has hardly
increased by a factor of ten. Thus, there is a gap between the amount of data
that can be processed and the amount of data that can be visualized. Large high-
resolution displays o er a way to deal with this gap and provide a signi cantly
increased screen area by combining the images of multiple smaller display devices.
The main objective of this dissertation is the development of new visualization and
interaction techniques for large high-resolution displays.
In the rst part of this thesis a hybrid display setup combining at panel display and
computer projector technology is presented to overcome the issue of discontinuous
images in multi-monitor systems. Discontinuities are primarily caused by monitor
frames and are the strongest argument against using highly parallel multi-monitor
systems. The approach eliminates these discontinuities and is thoroughly evaluated
in an extensive user study. Furthermore, a software framework is introduced that
enables developers to implement distributed applications for arbitrary display con-
gurations. Since large displays are usually driven by a cluster of render nodes, the
framework provides an abstraction layer to free developers from the complexity of
the underlying system when implementing applications.
The second part of the manuscript deals with interaction concepts for large high-
resolution displays. Due to their size many traditional interaction devices and
metaphors whose origins are in single display environments do not scale to the size
of large displays or to the size of user groups working with them. A new interaction
approach is introduced which is based on a combination of two-dimensional visual
tags and camera-enabled smart phones. The approach supports collaboration ofmultiple user groups in parallel and is implemented in two example scenarios on a
fty tile display cluster.
The third part of this dissertation deals with direct volume rendering on large high-
resolution displays and introduces a technique to visualize giga-scale data sets whose
sizes exceed the available, combined resources of all render nodes. The proposed out-
of-core technique makes e cient use of the resources available on modern graphics
cards and generates renderings of volumetric data sets larger than the texture bu er
size of a single graphics card at signi cantly higher levels of detail than on a single
desktop display. Furthermore, the thesis investigates how hierarchies of bounding
volumes can be exploited as acceleration structures for direct volume rendering to
achieve high frame rates during visualization.Zusammenfassung
Aufgrund technologischer Fortschritte hat sich die Leistungsf ahigkeit von Compu-
tersystemen in den letzten Jahrzehnten beachtlich gesteigert. Zieht man das Gesetz
von Moore zu Rate, so hat sich die Anzahl der Transistoren in integrierten Schal-
tungen etwa alle zwei Jahre verdoppelt, wobei der Trend sich fortzusetzen scheint.
Ahnliche Tendenzen lassen sich bei der Entwicklung der Speicherdichte, der Netz-
werkbandbreite und der Verarbeitungskapazit at ausmachen. Als Konsequenz ist die
Menge der verarbeitbaren Daten um viele Gr o enordnungen gestiegen. Gleichzeitig
allerdings hat sich das Au osungsverm ogen von Bildschirmen nur etwa verzehn-
facht. Folglich kla t eine Luc ke zwischen der Menge der Daten, die verarbeitet wer-
den k onnen und der Menge der Daten, die auf einem Bildschirm darstellbar sind.
Gro e hochau osende Displays bieten eine L osung, indem sie die Bilder mehrerer
kleinere Displays zu einem Gesamtbild vereinen. Der Schwerpunkt dieser Disserta-
tion liegt auf der Entwicklung neuer Visualisierungs- und Interaktionstechniken fur
gro e hochau osende Displays.
Im ersten Teil der Arbeit wird ein hybrider Flachbildschirm-/Projektor-Ansatz vor-
gestellt, um das Problem von Bilddiskontinuit aten in Multimonitorsystemen zu
osen.l Bilddiskontinuit aten werden prim ar durch Monitorrahmen hervorgerufen
und sind ein Hauptargument gegen den Einsatz hoch paralleler Multimonitorsys-
teme. Der vorgestellte Ansatz beseitigt die Diskontinuit aten und wird in einer
ausfuhrlic hen Benutzerstudie evaluiert. Des Weiteren wird ein Software-Framework
vorgestellt, welches Entwicklern die Implementierung verteilter Anwendungen fur
beliebige Kon gurationen von Displays erm oglicht. Da gro e Displays fur gew ohn-
lich mit Hilfe eines Clusters von Renderknoten betrieben werden, stellt das Frame-
work dem Entwickler eine Abstraktionschicht zu Verfugung, um ihn von der zu-
grunde liegenden Systemkomplexit at zu befreien.
Der zweite Teil des Manuskripts besch aftigt sich mit Interaktionskonzepten fur
gro e, hochau osende Displays. Viele Interaktionsger ate und Metaphern, deren Ur-
sprunge im Bereich der Einzelbildschirmsysteme liegen, lassen sich nicht ohne Wei-
teres hochskalieren, wenn die Gr o e der hochau osenden Displays oder die Anzahl
der Nutzer, die mit ihnen arbeitet, wachsen. Es wird ein neuer Interaktionsansatzpr asentiert, der auf einer Kombination von zweidimensionalen Bildmarkierungen
und sogenannten Smart Phones mit Kameraunterstutzung aufbaut. Der Ansatz
unterstutzt die gleichzeitige Zusammenarbeit mehrere Benutzergruppen. Er wurde
auf einem Displaycluster bestehend aus funfzig Monitoren implementiert, und seine
Anwendbarkeit wurde in zwei Beispielszenarien demonstriert.
Der dritte Teil der Dissertation besch aftigt sich mit der Volumenvisuali