Research

The parallel processing power of modern GPUs allows visualisation in almost any area to become interactive.  At VIS and VISUS, we are striving to make visualisation in all important areas like volume visualisation, particle visualisation, molecular surface visualisation, etc. faster and visually better. However, such new techniques often require the latest graphics cards, which might not be available to a wide user base. On the other side, networking technology nowadays provides users with an ubiquitous high-bandwidth access to the internet, even on mobile devices, which is sufficiently fast to stream live video. Our goal is to use and to improve such technology to make GPU-based high-performance visualisation available to all the users who need it and at an image quality a user in the field of imaging science can trust in.

Remote and distributed rendering has a long tradition at VISUS. We can build on our experience using GPU clusters, image generation for large high-resolution displays, remote rendering for mobile devices and hybrid rendering techniques using 3D capabilities available in all modern browsers. This will help bringing the latest visualisation technologies to all project partners, be it users on-site using our large-scale display devices or users working remotely on standard or even mobile devices.

Modern imaging technology, sensors and computer simulations not only provide us with an ever increasing amount of richer data, but also different kinds of views on a single problem. Different visualisation techniques exist for most of these modalities, e.g. volume rendering for scalar fields and glyph-based techniques for vector and tensor fields. Structured data, like information about patients and statistical data, require totally different approaches from the field of information visualisation. Our goal is to provide an integrated, holistic view which aggregates all the data in a single interface. This way, we want to enable seamless exploration of multi-scale, multi-field and multi-variate data.

We also intend to explore how different kinds of devices – ranging from standard desktop displays to large high-resolution displays, which can show a large amount in detail while providing its contextual information – can be used to address this problem and how visualisation must adapt to the display scenario. The visualisation facilities at the Imaging Science Centre will feature tiled displays using different technologies, which will enable comparing their applicability for real-world scenarios like collaborative analysis or teaching.

Research at VISUS covers the whole range of scientific visualisation, information visualisation, visual analytics, visual computing and computer graphics. Read more about all of these research fields at http://www.visus.uni-stuttgart.de/en/research.