Introduction

Previous visualisation research work at the Australia Telescope National Facility has concentrated on visual representation of large three-dimensional data sets to allow the astronomer to gain insights into the global structure of the data. See Norris (1994) and Gooch (1995abc) for more information. The next step in the process of extracting science from data involves obtaining quantitative information.

Something the astronomer would like to do is point at a feature of interest in the three-dimensional data cube, usually some small ``blob'' or ``lump'', and find out what the three-dimensional co-ordinate of the blob is.

The astronomer would also like to draw a three-dimensional polygon around a blob and do some further analysis on that sub-volume, such as computing the total flux.

Because of the nature of astronomical data, no sharply defined edges and surfaces appear solid. As a consequence, the impressive illusion of depth we are familiar with in other fields of visualisation (which often use surface-rendering techniques) tends not to be so dramatic when visualising astronomical data. We are left then with the problem of how to grasp the wispy tendrils of space without stabbing into the void.

To do this we must be able to determine the three-dimensional position of features of interest and display three-dimensional objects inside a volume of data. This paper discusses new research towards this goal.

Cursors

To determine the three-dimensional position of a feature the user sees, there must be some mechanism to point to the feature. This is analogous to the two-dimensional cursor (usually controlled by a ``mouse'') available in many image-display tools. Simply moving the cursor to a feature of interest and reading off the horizontal and vertical co-ordinates is not much help, because there is no information about depth. Is the feature close to the front or the back of the volume?

There are two main problems with implementing a three-dimensional cursor: how to input the position and how to display the cursor. This paper will focus on the latter problem.

Practicality

Astronomers generally consider themselves to be chronically under-funded, and generally do not have access to the latest computational resources. Many of the obvious ``brute-force'' techniques are not an option and hence algorithms to reduce computational effort are required. Some techniques in this area are also discussed.

© Copyright Astronomical Society of Australia 1997