Permalink
Browse files

changed slides based on Hussein's recommendations

  • Loading branch information...
timminata committed May 23, 2012
1 parent a3535e3 commit 8d4ddc7eaa03412bfadd2f4917ab1d510783370d
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
@@ -70,13 +70,14 @@ \subsection{}
\begin{columns}
\begin{column}{.65\linewidth}
\begin{itemize}
- \item Started by the UCT Department of Geomatics in 2004
- \item Aims to preserve cultural heritage
+ \item Started by the UCT Department of Geomatics
+ \item Aims to preserve cultural heritage by documenting heritage sites
+ \item Documenting a site involves producing:
\begin{itemize}
- \item Plans
- \item Laser scans
- \item Photos
- \item Videos
+ \item Detailed Laser scans
+ \item Photographs
+ \item Video Tours
+ \item Geographical Information Systems
\end{itemize}
\end{itemize}
\end{column}
@@ -92,9 +93,10 @@ \subsection{}
\begin{columns}
\begin{column}{.65\linewidth}
\begin{itemize}
- \item Over 100 models of sites in various African countries
- \item These models are big
- \item Containing upwards of a billion points
+ \item To create the model a point cloud is generated using a laser scanner
+ \item The project has produced close to 100 models of sites in various African countries
+ \item Some of the models are very large
+ \item Many contain over a billion points
\end{itemize}
\end{column}
\begin{column}{.35\linewidth}
@@ -105,11 +107,11 @@ \subsection{}
\end{frame}
-\begin{frame}{Problems Faced}
+\begin{frame}{Problems Faced by the Geomatics Department}
\begin{itemize}
\item Fast-growing volume of data
\item Difficulty in storing the data
- \item Difficulty with viewing the large models in real-time
+ \item Difficulty with viewing the large models
\item Data management issues
\item Data locality issues
\end{itemize}
@@ -118,7 +120,7 @@ \subsection{}
\begin{frame}{Problems Approached}
\begin{itemize}
- \item Difficulty with viewing the large models in real-time
+ \item Difficulty with viewing the large models
\item Data management issues
\end{itemize}
\end{frame}
@@ -147,7 +149,7 @@ \subsection{}
\section{Dynamic Point Cloud Streaming}
\subsection{Tim Trewartha}
-\begin{frame}{Dynamic Viewing of Large 3D Models}
+\begin{frame}{Dynamic Point Cloud Streaming}
\begin{itemize}
\item The Department of Geomatics has indicated that they have difficulties
handling the sizes of some of their models
@@ -158,36 +160,31 @@ \subsection{Tim Trewartha}
\end{itemize}
\end{frame}
+\begin{frame}{Current System}
+\includegraphics[scale=1.4]{images/laser.jpg}
+\end{frame}
-\begin{frame}{Research Question}
-\begin{itemize}
-\item Is it feasible to support real time viewing of models containing
- billions of points?
-\item Answering this question in the affirmative would enable exploration
- of the Zamani models in their full detail
-\item It would have a significant impact in the Geomatics department
-\end{itemize}
+\begin{frame}{Proposed System}
+\includegraphics[scale=1.4]{images/laser2.jpg}
\end{frame}
-\begin{frame}{Proposed Solution}
+\begin{frame}{Research Questions}
\begin{itemize}
-\item Implement a multi-resolution data structure to divide our model into
- manageable chunks
-\item Initially only a subset of the points is available
-\item As one zooms into the model and greater detail is required, we dynamically
- fetch additional points from our data structure until the full original
- detail is available
+\item Is it feasible to support viewing of models containing
+ billions of points at interactive frame rates?
+\item Can the level-of-detail data structure implemented facilitate streaming of
+ the models from a central server to clients?
+\item Answering these questions in the affirmative would enable exploration
+ of the Zamani models in their full detail
+\item It would have a significant impact in the Geomatics department
\end{itemize}
\end{frame}
-
\begin{frame}{Relevant Literature}
\begin{itemize}
-\item Common approaches to structuring large 3D models include octrees,
- R-trees, bounding sphere hierachies, and Hilbert Space Filling
- Curves
+\item Common approaches to structuring large 3D models include octrees and bounding sphere hierachies, as well as many others.
\end{itemize}
{\scriptsize
\begin{tabular}{l|c|l}
@@ -203,27 +200,43 @@ \subsection{Tim Trewartha}
\end{frame}
+\begin{frame}{Proposed Solution}
+
+\begin{itemize}
+\item Implement a multi-resolution data structure to divide our model into
+ manageable chunks
+\item Initially only a subset of the points is available
+\item As one zooms into the model and greater detail is required, we dynamically
+ fetch additional points from our data structure until the full original
+ detail is available
+\item The actual data can either be stored locally or streamed from a server
+\end{itemize}
+\end{frame}
-\begin{frame}{Proposed DataStructure}
+
+\begin{frame}{Data Structure}
\begin{itemize}
-\item From researching the literature it seems that octrees have the best
+\item From the literature it seems that octrees have the best
performance
\item All data is stored in the leaf nodes
\item Inner nodes provide simplified multi-resolution representations
\item No leaf node should contain more than a specified number of points
-\item Empirically, it seems that a value of around 30,000 gives good performance
+\item Empirically, it seems that a value of around 30,000 points gives good performance
\end{itemize}
\end{frame}
+\begin{frame}{Octree DataStructure}
+\includegraphics[scale=0.35]{oct.png}
+\end{frame}
+
\begin{frame}{Evaluation Criteria}
\begin{itemize}
-\item Can the system render the largest of the Zamani models at interactive
- frame rates?
+\item Has the hierarchical data structure successfully facilitated level-of-detail
+ streaming of the Zamani models?
+\item We will evaluate this by testing multiple models of varying sizes
\item If this goal is achieved the system will be a success
-\item Varying degrees of success can also be determined by testing smaller
- models of varying sizes
\end{itemize}
\end{frame}
}
@@ -266,7 +279,7 @@ \subsection{Michiel Baird}
\begin{itemize}
\item How effective is an automated workflow solution in Geomatics context?
\item Can data management be implemented using a workflow system?
-\item This would aid in reproducability of the research as well
+\item This would aid in reproducibility of the research as well
as address the data management issues
\end{itemize}
\end{frame}

0 comments on commit 8d4ddc7

Please sign in to comment.