Research Activities

Research Interests

My main interests are in image processing and sensor fusion applied to Automated Target Recognition (ATR) and Robotics. I am particularly interested in model based robust approaches to ATR using explicit physical models of the image formation process and multi-algorithms / multi-sensor fusion approaches. I am also interested in Adaptive Fusion Architectures enabling the classification process to be altered in real time depending on known and estimated algorithm performances and environmental parameters. I am also interested in autonomous navigation.

EPSRC Funded Research

I currently manage three EPSRC funded projects and two EPSRC studenships:

CONSENSUS: This project deals with concurrent sensor registration and fusion applied to multimodal image segmentation and classification. It is a 3 years program funded under the First Grant Scheme Initiative.

ATRIUM: This project looks at automated detection of concealed weapons using mm-wave imagery. At mm-wavelength, plastic and clothes are allowing the detection of weapons with very little illumination (unlike X Rays). This project is funded by EPSRC under the CRIME PREVENTION Initiative. It is a collaboration with QinetiQ, world leaders in mm-wave imagery.

Sonar Mine and Counter Measures: This project is the work of a PhD Student on Mine Detection using Sonar imagery. The technique is model-based and unsupervised and we have consistently obtained results of 80-90% detection and classification on different Seabed using different Sonars.

Student Autonomous Underwater Competition Europe: This project aims at creating a yearly European Autonomous Underwater Vehicle (AUV) challenge similar to the AUV competition currently run in the US/Canada. The challenge will be run for an initial period of 5 years and will be opened to any team comprising at least 75 % of students. Other team members could be professionals or academics.
After the revolution of the 1970’s, where remotely operated vehicles (ROVs) where first introduced amidst wide-spread scepticism from the off-shore industry, we are now seeing a new revolution, mainly driven by cost and deep-water exploration needs, where autonomous vehicles (i.e. no tether and support ship) are becoming the platform of choice for oceans survey, exploration, mapping, monitoring and surveillance. As a consequence, several commercial vehicles are now available.
However, whilst the engineering challenges have been mainly addressed, real autonomy is a largely unresolved problem. Solving it will require inputs from a new generation of highly motivated researchers and engineers. Fostering interest amongst engineering and computer science students and training them early will be the key to create the critical mass to allow real breakthrough. This will have a broader impact on autonomous vehicles as a whole (UAVs, Space Landers) as they share common problems.
Currently only a handful of expensive vehicles requiring a support ship and large infrastructure are available in Europe. Building small scale, inexpensive vehicles will enable research to be carried out at low cost and provide the necessary resources for complex multiple vehicle work.

EU Funded Research

I am also involved in several other EU funded projects:

AUTOTRACKER: This EU funded project (Framework V) is developing technologies for autonomous survey of pipelines and cables for the off-shore industry. I managed the early stages of the project and supervise the sonar tracking and obstacle avoidance algorithms development.

AMASON: AMASON (Advanced MApping with SONar and Video) provides marine science with an inexpensive, modular, reconfigurable multi-sonar and video sensor system, with advanced data processing algorithms implemented within a geographic information system (GIS). I am supervising the sensor fusion and sonar processing aspects of the project.

Industrial Research

I am involved in Knowledge Transfer Partnerships programs with Optos Ltd, a compay developing high resolution imaging devices for eye care. We help them develop automated screening methods based on image processing.

I am also the Chief Technical Officer of SeeByte Ltd, a spin out of the Oceans Systems Laboratory.