Volume rendering systems are designed to provide means to enable scientists and
a variety of experts to interactively explore volume data through 3D views of the
volume. However, volume rendering techniques are computationally intensive tasks.
Moreover, parallel distributed volume rendering systems and multi-threading architectures
were suggested as natural solutions to provide an acceptable volume rendering
performance for very large volume data sizes, such as Electron Microscopy data
(EM). This in turn adds another level of complexity when developing and manipulating
volume rendering systems.
Given that distributed parallel volume rendering systems are among the most
complex systems to develop, trace and debug, it is obvious that traditional debugging
tools do not provide enough support. As a consequence, there is a great demand to
provide tools that are able to facilitate the manipulation of such systems. This can be
achieved by utilizing the power of compute graphics in designing visual representations
that reflect how the system works and that visualize the current performance state
of the system.The work presented is categorized within the field of software Visualization, where
Visualization is used to serve visualizing and understanding various software. In
this thesis, a number of visual representations that reflect a number of functionality
and performance aspects of the distributed HVR, a high quality volume renderer
system that uses various techniques to visualize large volume sizes interactively. This
work is provided to visualize different stages of the parallel volume rendering pipeline
of HVR. This is along with means of performance analysis through a number of
flexible and dynamic visualizations that reflect the current state of the system and
enables manipulation of them at runtime. Those visualization are aimed to facilitate
debugging, understanding and analyzing the distributed HVR.
| Date of Award | Jul 2012 |
|---|
| Original language | English (US) |
|---|
| Awarding Institution | - Computer, Electrical and Mathematical Sciences and Engineering
|
|---|
| Supervisor | Markus Hadwiger (Supervisor) |
|---|