Research

I am a postgraduate student under the supervision of Dr Peter Kovesi in the Vision and Visualisation Research Group. My research lies in the field of Computer Vision, particularly the analysis of motion fields for discerning scene information. Currently I am investigating how to extract the divergence, curl, and deformation of optical flow fields in the face of noise. These first-order differential quantities are useful for the recovery of 3D surface normals and determining time-to-contact in ego-motion calculations.

Some reading

Past work

During my Honours year, I worked with Professor George Milne on reproducing the dynamics of disease spread using a cellular automata (CA) model.

I have also worked on a project examining the use of evolutionary algorithms for tertiary protein structure prediction, pursuing the hypothesis that protein folding can be modelled as an optimization process. Assuming that a protein's final conformation is a compromise between factors such as potential energy, orientation, and hydrophobia; a multi-objective evolutionary algorithm can provide an estimate of that conformation. As a prototyping exercise, I investigated the use of a multi-objective evolution strategy for the problem of automated map labelling.

Back Home


Last modified: Thu Nov 17 09:27:25 W. Australia Standard Time 2005