Dr Peter Yuen

Peter Yuen

Senior lecturer
Location: WH132
E: P.yuen@cranfield.ac.uk
T: +44 (0)1793 78 5205
Informatics and Systems Engineering

Current activities

Machine Vision Research
Spectral Vision for automatic target recognition

PhD Programmes in :

data mining
data compression
remote sensing
long range target detection
scene calibration
spectral constancy
machine learning
pattern recognition
automatic target recognition
target tracking
biometric spoofing
detection of intent
scene understanding
real time processing

Clients

Background

Started his research career in 1985 at Imperial College, London in semiconductor physics concerned with molecular beam epitaxy growth, band structure modelling and device characterisation
In 1997 joined the corporate laboratory of Asahi Chemical Industry Ltd as a chief scientist, and explored how surface acoustic wave (SAW) devices could be applied to mobile communication applications. During this time he filed a Japanese patent application on how semiconductor thin film could be grown on a piezoelectric substrate in order to amplify Surface Acoustic Waves using a field effect configuration
In 1999 he joined DERA, which is now known as QinetiQ, as a principal scientist (level 7) and engaged in defence research specifically for the detection of submarines in littoral water. He proposed a dual approach using a stochastic neural network and statistical pencil matrix methods for un-mixing weak signals blindly from a strong background. The novelty of the proposal secured 3-years funding under the MOD’s Fellowship CRP programme
In 2002 hewas elected a Fellow of the Institute of Physics and the Institute of Mathematics and Its Applications
In late 2003 he moved on and accepted a senior principal scientist post (level 8) with BAE Systems, Advanced Technology Centre, Optics and Laser Technology section in Filton, to deal with hyperspectral research. During this period he published a number of fundamental and applied research papers related to target detection phenomena especially for high dimensional hyperspectral data
In May 2007 heaccepted a senior lecturer post with Cranfield University at the Defence Academy of the united kingdom and initiated a programme of Spectral Vision research for military and homeland security applications
He is active in a new stream of machine vision research, namely Spectral Vision, for automatic target recognition. His research is aimed at homeland security and military applications, he has acquired a sophisticated hyperspectral camera system for his research
At present, one of his PhD students is investigating data mining aspects specifically for finding how high dimensional data can be classified efficiently and accurately even when the data clusters are close to each other.

PhD Students

T Chen - Remote assessment of intent by hyperspectral sensing of human physiological responses triggered by stress
K Hong - Exploitation of spectral and textural information for the detection of anomalous behaviour in a cluttered background
H Kam - Assessment of improvements to classification accuracy that can be achieved by an adaptive feature selection scheme
A Tsitiridis - Object recognition and tracking using cortex like machine vision processing