MS Thesis

Home » Academia » RIT » MS Thesis

My Master’s Thesis was started in February 2006 and defended on June 21, 2006. The abstract and all deliverables are available below as well as an archive of updates that were posted on this site.

Abstract

N-body simulations solve the n-body problem numerically and determine the trajectories of the n point masses. The result of these calculations is a huge amount of data (up to tens of gigabytes) detailing the positions and other properties of each body such as mass and velocity. To effectively draw conclusions from this data, one must employ scientific visualization to create images and movies that illustrate the structure of the data. This thesis seeks to apply the computer animation technique of skeletonization to the volume data produced by n-body simulations in order to extract the inner structure of the data. This novel application will be compared to traditional point and volume rendering methods in terms of their ability to visualize astrophysical phenomena hypothesized to be present in the data. All of the techniques will be implemented in Java for the Spiegel visualization framework.

Files

BibTex Citation

@mastersthesis {edwarddale-2006-rit,
   author = "Edward Dale",
   title = "Visualizing the Inner Structure of N-Body Data using Splatting and Skeletonization",
   school = "Rochester Institute of Technology",
   month = "June",
   year = "2006"
   abstract = "N-body simulations solve the n-body problem numerically and
      determine the trajectories of the n point masses.  The result of these
      calculations is a huge amount of data (up to tens of gigabytes) detailing
      the positions and other properties of each body such as mass and velocity.
      To effectively draw conclusions from this data, one must employ scientific
      visualization to create images and movies that illustrate the structure of
      the data.  This thesis seeks to apply the computer animation technique of
      skeletonization to the volume data produced by n-body simulations in
      order to extract the inner structure of the data.  This novel application
      will be compared to traditional point and volume rendering methods in
      terms of their ability to visualize astrophysical phenomena hypothesized
      to be present in the data.  All of the techniques will be implemented in
      Java for the Spiegel visualization framework.",
}