Information Interaction: Improving Visualization Methods
Efficient extraction of useful knowledge from large and complex data requires powerful means to enable the user to perceive the data and its structure. Visualization techniques, presenting information to the user through the high-bandwidth senses of vision and touch, can provide a means by which new understanding can be gained quickly and efficiently. The importance of interactivity in the interpretation process, to enable easy location, extraction and display of the pertinent data, cannot be over-estimated. This places heavy demands not only on information representations but also on data management, processing and interaction mechanisms in the visualization pipeline. Specific research issues for MOVIII are:
- Large Scale Data in Adaptive Visualization Pipelines: Our previous work has shown that improved volume rendering pipelines can be obtained using the domain knowledge encoded in transfer functions. In MOVIII we will make use of other advanced modeling techniques, combined with expertise from the MOVIII consortium, to improve upon these schemes and so enable interactive analysis of multi-dimensional, multivariate data in next generation visualization pipelines.
- Interactive Environments for Visual Decision Support: We have had considerable success in the development of new visualization and interaction methods and have found that, for complex time-based data, these approaches can provide great insight into model fitting and so guide the selection of appropriate models and parameters. This approach has proven extremely useful in system identification. In this project we will address new 3D and 4D interaction mechanisms, based on VR technologies, for the visualization of dynamic knowledge. Our target application here is decision support in flight control. Previous work has focussed on the management of flight data but, within this project, we will focus on the control of multiple unmanned air vehicles, and the interpretation of data retrieved from them.
- Multimodal Interaction Techniques: Recent research has shown the benefit of multi-sensory information representations for multivariate and complex data. The NVIS research group has developed haptic algorithms for static volumetric data to enable the interpretation of features which are difficult to display graphically or by providing additional information to support the visual sense. We have had success with these techniques across application areas from engineering to medical diagnostics and surgical planning. In the context of the MOVIII pro ject we will extend this haptic interaction to time-varying and deformable volume data using models of the data which respond to user actions according to material properties and forces applied.