Visualisierung 2 - Streamline Variability Plots

Nicolas Grossmann, 1325103
Thomas Köppel, 1327052

Overview

Clustering Result of Streamline Data

We have implemented a flow visualization tool after "Streamline Variability Plots for Characterizing the Uncertainty in Vector Field Ensembles" by Ferstl et al. 2016 [1] allowing stream- and pathlines to be clustered into different trends emphazising a streamline median (middle line of a cluster), the convidence lobes around the median and the number of streamlines in a specific cluster by a coloured barplot. The streamlines are firstly generated, pca is transforming the data allowing clustering. In each cluster a streamline median is determined and streamlines are sampled back and plotted in 2D using splatting and surrounded by a boundary. These Streamline Variability Plots show trends and statistical parameters (like the confidence interval) in the same domain the data resides.

Furthermore, we applied this technique to a new kind of data: Travel Data, to create an informative visualization showing major tourism trends.

Clustering Result of Travel Data

Configuration

How to use

The typical workflow of the program consists of 3 steps, Load, Simulate and Cluster, by clicking on the respective buttons on the right panel.

Documentation

C Sharp Documentation
Matlab Documentation

Download

Source
Binary

Reference

[1] Ferstl, Florian, Kai Bürger, and Rüdiger Westermann. "Streamline variability plots for characterizing the uncertainty in vector field ensembles." IEEE Transactions on Visualization and Computer Graphics 22.1 (2016): 767-776.