Digital Commons Network^™

Isosurface Extraction In The Visualization Toolkit Using The Extrema Skeleton Algorithm, Subha Parvathy Mahaadevan

Masters Theses

Generating isosurfaces is a very useful technique in data visualization for understanding the distribution of scalar data. Often, when the size of the data set is really large, as in the case with data produced by medical imaging applications, engineering simulations or geographic information systems applications, the use of traditional methods like marching cubes makes repeated generation of isosurfaces a very time consuming task. This thesis investigated the use of the Extrema Skeleton algorithm to speed up repeated isosurface generation in the visualization package, Visualization Toolkit (VTK). The objective was to reduce the number of non-isosurface cells visited to generate …

Itera- Tive Reconstruction Framework For High-Resolution X-Ray Ct Data, Thomas Matthew Benson

Masters Theses

Small animal medical imaging has become an important tool for researchers as it allows noninvasively screening animal models for pathologies as well as monitoring dis- ease progression and therapy response. Currently, clinical CT scanners typically use a Filtered Backprojection (FBP) based method for image reconstruction. This algorithm is fast and generally produces acceptable results, but has several drawbacks. Firstly, it is based upon line integrals, which do not accurately describe the process of X-ray attenuation. Secondly, noise in the projection data is not properly modeled with FBP. On the other hand, iterative algorithms allow the integration of more complicated sys- …

3d Visualization Modules For Chemical Engineering – A Web-Based Approach Using Java And Opengl, Sharad Anant Gupta

Masters Theses

The main objective of this work is to implement web-based educational modules for chemical engineering students. Phase behavior is a topic with which the students seem to struggle with, particularly for mixtures, where a 2-D representation of the phase diagram falls far short of the understanding a 3-D model can provide. Using the platform-independence of Java and the graphics capability of OpenGL, three phase diagram Java applets have been developed. Users can view these web-based 3D applets by installing a plug-in. These modules provide users with an ability to rotate the 3D models, slice through them, zoom into them and …

Data Access In Wide Area Networks Of Heterogeneous Workstations, Kim Buckner

Doctoral Dissertations

The accessibility of data in wide area networks can be difficult. This research shows the use of the Internet Backplane Protocol (IBP) along with a modified version of the C standard I/O library that can allow data to be easily accessible without having to make major modifications to legacy code. In fact if legacy programs only use standard input and output routines, they need only be recompiled to effect a homogeneous file system. It also demonstrates that this access is predictable enough to make decisions on what data to access and in what fashion that access is most effective.

Integrated Visualization Of Diffusion Tensor And Functional Mri, Nathaniel Richard Fout

Masters Theses

To understand the morphology, structure, and function of the human brain and the underlying relationships therein has long been a goal of mankind. Technologies are constantly emerging and evolving in an effort to realize this goal, with each new development potentially providing another piece of the puzzle. Among those technologies are advanced imaging modalities such as Diffusion Tensor MRI (DT-MRI) [2] and Functional MRI (fMRI) [8], whose purpose is to provide deeper insight into the neural network connecting functional units of the cerebral cortex. More specifically, DT-MRI captures a description of the fibrous structures (such as nerve fibers) in the …

A Preemption-Based Meta-Scheduling System For Distributed Computing, Sathish Vadhiyar

Doctoral Dissertations

This research aims at designing and building a scheduling framework for distributed computing systems with the primary objectives of providing fast response times to the users, delivering high system throughput and accommodating maximum number of applications into the systems. The author claims that the above mentioned objectives are the most important objectives for scheduling in recent distributed computing systems, especially Grid computing environments.

In order to achieve the objectives of the scheduling framework, the scheduler employs arbitration of application-level schedules and preemption of executing jobs under certain conditions. In application-level scheduling, the user develops a schedule for his application using …

Performance Improvements Of Common Sparse Numerical Linear Algebra Computations, Piotr Rafal Luszczek

Doctoral Dissertations

Manufacturers of computer hardware are able to continuously sustain an unprecedented pace of progress in computing speed of their products, partially due to increased clock rates but also because of ever more complicated chip designs. With new processor families appearing every few years, it is increasingly harder to achieve high performance rates in sparse matrix computations. This research proposes new methods for sparse matrix factorizations and applies in an iterative code generalizations of known concepts from related disciplines. The proposed solutions and extensions are implemented in ways that tend to deliver efficiency while retaining ease of use of existing solutions. …

Sparta: A Graphical User Interface For Malicious Mobile Code Fingerprint-Ing., Victor P. De Cerqueira

Masters Theses

This thesis introduces and describes SPARTA (for Stochastic Profiling Application for the Rendering of Trees and Automata), a graphical user interface used as a front end to a collection of tools written in C that collectively convert a log of registry system calls performed by an application into binary descriptions of PSTs (for Probabilistic Suffix Trees) and PSAs (for Probabilistic Suffix Automata), which are models used to represent application behavior on Windows-based systems. SPARTA works by rendering these binary descriptions into graphical form, showcasing a variety of features intended to make the user interaction with PSTs and PSAs informative and …

Thermal Evolution Of Planetesimals And Protoplanets In The Terrestrial Planet Region: Code Optimization And Implementation On A Distributed Grid Using Netsolve, Amitabha Ghosh

Masters Theses

A code for asteroidal heat transfer and growth is optimized for performance. The Gauss elimination routine for the solver is replaced by a sparse matrix routine. Finite element matrix assembly operations are rewritten to reduce operations involving 3D arrays to 1D. Advantage is taken of the sparse matrix structure of finite element matrices in reducing 2D arrays to 1D. The number of vector touches are reduced to the extent possible, by carrying over statements from one iteration to the next. The number of do loops are reduced by merging several do loops into one. The optimization reduced the CPU time …

A Coarse-Grain Parallel Implementation Of The Block Tridiagonal Divide And Conquer Algorithm For Symmetric Eigenproblems., Robert M. Day

Masters Theses

Cuppen’s divide and conquer technique for symmetric tridiagonal eigenproblems, along with Gu and Eisenstat’s modification for improvement of the eigenvector computation, has yielded a stable, efficient, and widely-used algorithm. This algorithm has now been extended to a larger class of matrices, namely symmetric block tridiagonal eigenproblems. The Block Tridiagonal Divide and Conquer algorithm has shown several characteristics that make it suitable for a number of applications, such as the Self-Consistent-Field procedure in quantum chemistry.

This thesis discusses the steps taken to implement a coarse-grain parallel version of the Block Tridiagonal Divide and Conquer algorithm, suitable for a parallel supercomputer or …