Integral surfaces, such as stream- and path-surfaces, are highly effective when it comes to the exploration and analysis of the long-term behavior of three-dimensional flows. However, specifying the seeding curves that lead to an expressive set of integral surfaces is a challenging and cumbersome task. In this paper, we propose an algorithm for automatically seeding multiple stream-surfaces around a user-specified point of interest. The process is guided by a streamline similarity measure. Within the resulting integral surfaces, adjacent streamlines are as similar as possible to each other. Conversely, we aim at conveying different aspects of the flow behavior with each surface. This is achieved by maximizing the dissimilarity between streamlines from different stream-surfaces. The capabilities of our technique are demonstrated on a number of application cases. We compare our expectations with the results of an informal survey. We report from our detailed exchange with a domain expert concerning the expressiveness and usefulness of our approach. A thorough analysis of the few parameters involved is provided.

Supervisor: Helwig Hauser
Project Report: Here

Some screenshots from the developed prototype: