 |
|||||||||||||||
|
|
||||||||||||||
|
|
||||
SimVis ModellingThe SimVis Modelling theme is structured into three project clusters, which reflect the various phases in the construction, production and usage of visual representations. Modelling begins with the model-building phase, during which an abstract representation of the data or scene of interest is generated. This in turn is submitted to the rendering phase, ready for interpretation by the user via some appropriate display medium. This latter might, for instance, be the wall-sized display of the HIVE, the haptic feedback device known as the Reachin™, or the highly immersive Elumens VisionStation® wrap-around screen.
Model-building Within the model-building cluster, the group is researching the underlying
mathematics of implicit curve and surface design [1] and smooth blending
of the shapes that are a key part of computer graphics. Physically based
modelling, on the other hand, allows the creation of virtual environments
that automatically compute the motion of objects, both soft and rigid,
based upon a detailed physical model [2,3]. In marine visualization
the emphasis is on bringing together a number of different data streams,
such as a depth profile acquired in real-time via sonar and a CAD (Computer
Aided Design) model of some underwater obstruction obtained previously
[4,5]. RenderingThe rendering cluster concentrates on producing the actual graphics,
bearing in mind the way the model has been built, the computer power
available, and the application’s intended use. Thus for architectural
models that have been acquired using our Leica Geosystems laser scanner,
point-rendering and hybrid point-polygon rendering methods are being
investigated. Our cartographic work involves a branch of graphics known
as non-photorealistic rendering, in this case to produce artist-like
sketches that have nonetheless been drawn by the computer [6]. Besides
rendering the graphics of the model we are also researching the rendering
of haptics and audio for a model. Haptics allows a user to touch and
feel a model; this can be achieved using the Reachin™ which displays
co-registered graphics and haptics. Our haptics work includes high fidelity
rendering of frictional forces for contacting surfaces [7]. InterpretationThe interpretation cluster is perhaps the most diverse of the three and can best be described as interaction with data, simulations or real-life scenes. Think of a weather map–how much more difficult would it be to understand where the wind was blowing from if it were shown as a speed and heading, rather than the more familiar arrows? This is scientific visualization, where the numbers in data are given a graphical representation in order to try to understand what is going on. Data is also at the heart of our work on multidimensional and multivariate visualization, and both these projects are looking at new ways to use the various different display systems mentioned above. Data might itself be generated by a simulation–in the weather example this would be the forecasting software used by the meteorologists–but as such it is only a snapshot to be looked at after the computation has ended. In computational steering, however, the simulation itself is brought into play [8]. Using this approach, scientists can carry out experiments on the computer in much the same way as they used to in a laboratory using test tubes and Bunsen burners. Finally there is interaction with real-life scenes, commonly termed ‘Virtual Reality’. In collaboration with the Psychology department, we have developed software that allows such scenes to be presented under different conditions of perspective, lighting, shading, motion parallax and with stereoscopic display. This software has recently been used to study the influence of stereoscopic vision in a face recognition task [9]. Virtual reality likewise finds application in a host of other areas, including architecture: our work on the St Stephens initiative in Hull raised awareness of this important project before the site had been fully cleared of its old and outdated structures.
[1] Implicit Fitting Using Radial Basis Functions with
Ellipsoid Constraint, Li Q, Wills DPM, Phillips R, Viant WJ, Griffiths
JG, and Ward JW, Computer Graphics Forum, 23(1), pp 1-15, 2004. [9] Transfer between 2D and 3D Representation of Faces,
Liu CH, Ward JW, Young AW. (to appear) Visual Cognition. [Full SimVis Publication lists here]
|
||||
|
||||