Visualisation Overview

What is visualisation?

Scientific visualisation is, in essence, the process of converting numerical data into a visual format which facilitates the interpretation of those data.

A considerable portion of the processing power of the human brain is related to visual cognitive processes; that is, the interpretation of visual stimuli and its role in allowing us to move aound in, survive, manipulate and appreciate the physical world around us.

Visual clues, such as colour, form, perspective and scale, give us much information as to the nature of objects in the physical world, and our relationship to those objects. Thus, the visual interpretation of data- when those data are presented as objects with properties we can directly sense- is a very powerful method for gaining understanding of those data.

Data may be visualised graphically- showing the behavioural relationship between two or more variable quantities in a manner that allows us to see how one quantity affects the other. This is, perhaps, the oldest form of data visualisation, and should be familiar to anyone who has studied basic mathematics.

However, when we start dealing with large volumes of data, with multiple variables, and with data that change with more than one variable at a time, it is better to look at three-dimensional (3D) visualisation of those data. This is where modern computional techniques come to the fore.

3D Visualisation

As mentioned above, by virtue of the way we interpret the physical world around us, somehow presenting data as objects which can be viewed, rotated, animated over time, perhaps even felt, allows us to see what's going on in the data much more readily than viewing a list of thousands or millions of numbers.

Allied to the development of fast and powerful computer systems, is the ability to model entire physical systems, and to review and predict the behaviour of those systems over time. This is very often done in three spatial dimensions (though this is not always necessary, or even feasible...) and over time, and the output can consist of millions of numbers.

Visualisation software allows such data to be displayed as a three-dimensional scene, showing the values of the variables at each modelled spatial and temporal location; this permits a very intuitive connection between the visualised scene (and thereby its mathematical basis)and the true, physical system which has been modelled.

The software uses visual concepts such as colour, transparency, perspectives, scaling, direction, etc, in order to represent various aspects of the variable quantities in the model; these correspond to different colours, etc, in the visualised scene.

A certain amount ot interactivity is usually present; the viewer can rotate the scene, zoom in or out, change some of the display parameters, for example, in order better to view the visualised data.In some cases, it is possible for the viewer to "steer" the computations in real time, so that changes to the model conditions and their results may be seen in real time.

Visualisation can also be used for the displaying of models of physical objects; CAD/CAM designs for engineering and architecture, for example.

Stereo visualisation

Whilst depth cues (shading, fogs, hidden lines, etc) can be used to give the illusion of distance to 3D visualisations presented on a flat display, a greater realism can be obtained via stereoscopic displays.

In these, each eye is given a separate image (either still or video), each rendered from a slightly different point of view. The brain will combine these and interpret the view as being that of a fully three-dimensional, "solid" scene.

A number of methods are employed to do this, amongst them anaglyphic images (e.g. red/green stereo images), polarised lens systems and the Infitec filtered projection system. Recent products allow a stereoscopic computer monitor display with no extra hardware.

Visualisation environments

In many cases, the effectiveness of visualisation is improved by, as much as possible, immersing the viewer within the visualisation. This is most commonly achieved by projecting the visualistion images on a screen of sufficient size to fill most of the viewer's visual field.

This can involve monoscopic or stereoscopic projection onto a single screen in a darkened room, or similar projection onto curved or multiple screens such that the scene more than fills a static viewpoint, and looking or moving around is required to see all of the scene.

Extra interaction and sensory immersion is allowed by the use of motion-tracking equipment in order to change the viewpoint as the viewer moves, or force-feedback ("haptic") devices to allow a certain amount of tactile interaction with objects in the scene.

SAPAC facilities

SAPAC VisLab

SAVRC (South Australian Virtual Reality Centre)