Biological Shape Measurement
 
Noncontact 3D measurement In 1997, as a Research Fellow in the Departments of Computer Science and Biochemistry at the University of Western Australia, I was part of a major, ongoing study into Human Lactation, involving noncontact 3D breast measurement. Some aspects of this project are reviewed here.

The figure to the left illustrates a simple measurement taken of a particular subjects head, showing the actual measured points, as well as the resulting texture mapped model from two different viewpoints.

The problem of occlusion is quite clear from this figure - there is no data from any areas invisible to the camera, such as much of the side of the head, under the chin, and even the side of the nose. To help alleviate this problem, I worked on enhancing the basic system to employ three independent camera/projector pairs, and thus integrate data from three different viewpoints.

Human Lactation I also worked on integrating into the system thermographic data obtained using an infrared camera. The inclusion of temperature information was particularly interesting since it gave an indication of the underlying synthetic activity, as shown in the images below, one of a mother lactating in her right breast only, and one of a mother lactating in her left breast only. Note the large temperature difference between breasts in both cases, with the lactating breast as much as 4 degrees warmer than the other.

Lactation in right breast only Lactation in left breast only
Thermal Image Thermal Image

Thermography The thermography study also clearly showed the drop in temperature during breastfeeding as illustrated in the following figure. The six images shown were taken at regular intervals during a single feed, and the temperature drops from around 35 degrees in the first picture (top left frame), to approximately 30 degrees in the final picture (bottom right frame).

Macintosh port of SHAPE Another application of the noncontact 3D measurement system I was involved in was in 1995 as part of a project for Fremantle Hospital in Perth. This work involved porting the system to the Macintosh platform, and incorporating colour information in the image analysis, in order to construct a system that could be used to aid in the diagnosis and treatment of leg ulcers.

Selected Papers
  • K.P. Wessen, A portable 3D shape measurement system, Second Asian Conference on Computer Vision, Nanyang Technological University, Singapore, III:pp 695-699, 6-8 December, 1995.
  • R.J. Sadleir, K. Wessen, M. Cheng, R. Owens, and P.E. Hartmann, A three view shape reconstruction system for soft tissue volume estimation: interface and implementation, in C. MacNish, A. Czarn, and P. Taylor, editors, The Ninth University of Western Australia Computer Science Research Conference, volume 1, pages 203-204, Mandurah, Western Australia, 21-22 September 1998.

Computational Fluid Dynamics
 
3D Hydrodynamic and Water Quality Modelling

In 1996 I was a Research Fellow in the Centre for Water Research at University of Western Australia. While there, I designed and built ELMO, a three dimensional Estuary and Lake Model. ELMO, written in Fortran 90, employed a semi-implicit discretisation of the Navier-Stokes equations, assumed hydrostatic vertical pressure, and used a Euler-Lagrangian advection scheme. A turbulent closure mechanism was included, and the model joined to a three-dimensional water quality model.

Click here or on the image below to download a movie produced from an ELMO simulation of Lake Kinneret in Israel (2.5 MB).
 

Selected Paper
  • M.P. Bonnet and K.P. Wessen, ELMO, a 3-D water quality model for nutrients and chlorophyll: first application on a lacustrine ecosystem, Ecological Modelling 141, 19-33, 2001.

Theoretical Plasma Physics
 
Nonlinear Dynamics

From 1990 to 1993 I was a PhD student in the Department of Theoretical Physics and the Plasma Research Laboratory, Research School of Physical Sciences and Engineering, at the Australian National University. My thesis research was in the nonlinear dynamics of plasma instabilities, specifically tearing modes and collisional drift waves. I looked at the Invariant Manifold Reduction, which is a generalisation of the Centre Manifold Reduction, and allows the construction of a low-dimensional description of the long-term dynamics of an infinite dimensional nonlinear system.

The two Quicktime movies below show animations of two interesting and stable nonlinear states of the tearing mode on a 4D invariant manifold. The first shows the development of a travelling wave, and the second a stable heteroclinic cycle where two unstable m=2 modes are linked to form a stable cycle switching back and forth between them.

Travelling wave (870KB) Stable heteroclinic cycle (1.4MB)

Magnetohydrodynamics My first taste of research was as a 4th year student in Theoretical Plasma Physics in the School of Physics at Sydney University in 1989, studying surface waves on current sheets. This lead to studying tearing modes in rotating plasmas for the first six months of my PhD, before becoming more focussed on the nonlinear work described above.

Selected Papers
  • K.P. Wessen and M. Persson, Tearing mode stability of a cylindrical plasma with equilibrium flows, Journal of Plasma Physics 45, 267-283, 1991.
  • K.P. Wessen and N.F. Cramer, Finite-frequency surface waves on current sheets, Journal of Plasma Physics 45, 389-406, 1991.
  • K.P. Wessen, Tearing Mode Evolution on an Invariant Manifold, Physics of Plasmas 2(2), 370-387, 1995.
  • K.P. Wessen, Application of the Invariant Manifold Reduction to Dissipative Plasma Instabilities, Physics of Plasmas 2(2), 586-588, 1995.