Navigation in time-varying scientific data

Wolter, Marc; Kuhlen, Torsten (Thesis advisor)

Aachen / Publikationsserver der RWTH Aachen University (2010) [Dissertation / PhD Thesis]

Page(s): VI, 167 S. : Ill., graph. Darst.

Abstract

The temporal resolution of today's simulated scientific data sets is constantly growing. Simulations spanning thousands of discrete time steps are common these days; simulations with a much higher resolution exist. The growing amount of discrete time steps leads to significant problems, particularly concerning the navigation through such data sets. This thesis presents a set of techniques that together enable an intuitive, interactive, and accurate navigation through time-varying scientific data. A time model that describes different time scales and their relations in scientific visualizations serves as a basis. Novel 3D interaction techniques for search and maneuver navigation tasks are developed and are evaluated in user studies. Based on initial user input, the temporal resolution of time-varying scientific data is adapted by a multi-objective optimization approach, which tries to combine conflicting analysis goals. In order to speed up the computation of time-varying visualizations during navigation, a remote parallel computation system using different user-centered scheduling strategies is presented. The developed approaches are evaluated using multiple large time-varying data sets from medicine, geosciences and flow mechanics. The heterogeneous techniques proposed - applied individually as well as in combination - enable or ease the analysis of large time-varying data in an interactive work process.

Identifier

  • URN: urn:nbn:de:hbz:82-opus-31942
  • REPORT NUMBER: RWTH-CONV-113761