Research projects at Universität Stuttgart
Originally, scatterplots are meant to show discrete data samples. In this project, we refined the idea of scatterplots to allow the visualization of continuous data. More information about the project can be found on the project page.
Animation of Orthogonal Texture Patterns for Vector Field Visualization
This project is an extension to our EuroVis 2007 paper. The original paper received an invitation for IEEE Transactions on Visualization and Computer Graphics (TVCG). In the new version of this paper, the original concept of orthogonal vector field visualization is further improved. More information can be found on the project page.
Academic Projects at Simon Fraser University
Volume Rendering with Higher-Order Interpolation
This project was done as part of the SFU CMPT 775 ("Visualization") course offered by Daniel Weiskopf during the Spring 2007 semester. The goal of the project was to implement higher-order interpolation methods in order to improve the visual quality of volume rendering applications. The challenge is to maintain a high rendering speed that lies in the range of interactive frame rates. More details are available on the course webpage.
Desktop Workload Evaluation
For SFU CMPT 886 ("System Software Design for Chip Multithreaded Processors") of Alexandra Fedorova the workload of desktop systems was analyzed. The approach and outcomings of this project are described in the corresponding conference paper (which is available for download on the project page).
Vector Field Visualization
For the SFU CMPT 770 ("Advanced Computer Graphics") course of Daniel Weiskopf, a program was developed which was the basis for a conference paper. The course project has also its own course webpage where you can basically find the same information as in the paper, but in a more compact fashion.
The second course project in the Fall 2006 semester was done for the SFU CMPT 882 ("Computer Vision") course offered by Greg Mori. At first, this project did not include a real research component - I just reimplemented a technique that was published earlier in a conference paper. The goal of this course project was to recognize actions of people in surveillance videos. The basic idea is to compare a reference "motion descriptor" with the motion descriptors of the persons in the vidoes. This is done in a frame-by-frame fashion by a set of MATLAB scripts that I created during the project. As it turned out, the analyzed motions provide a good data set for the action visualization framework of Ralf Botchen. This work is described in the paper "Action-Based Multifield Video Visualization" that will appear in IEEE Transactions on Visualization and Computer Graphics, Vol. 14, No. 4, July/August 2008.
Last modified: Nov. 5th, 2007
by Sven Bachthaler