Importance-Driven Particle Techniques for Flow Visualization

摘要

Particle tracing has been established as a powerful visualization technique to show the dynamics of 3D flows. Particle tracing in 3D, however, quickly overextends the viewer due to the massive amount of visual information that is typically produced by this technique. In this paper, we present strategies to reduce this amount at the same time revealing important structures in the flow. As an importance measure, we introduce a simple, yet effective clustering approach for vector fields, and we use scalar flow quantities at different scales in combination with user-defined regions of interest. These measures are used to control the shape, the appearance, and the density of particles in such a way that the user can focus on the dynamics in important regions at the same time preserving context information. We also introduce a new focus for particle tracing, so called anchor lines. Anchor lines are used to analyze local flow features by visualizing how much particles separate over time and how long it takes until they have separated to a fixed distance. It is of particular interest if the finite time Lyapunov exponent - a scalar quantity that measures the rate of separation of infinitesimally close particles in the flow - is used to guide the placement of anchor lines. The effectiveness of our approaches for the visualization of 3D flow fields is validated using synthetic fields as well as real simulation data.