Computer vision and cognitive systems computing

This research area aims to develop and apply algorithms for interpreting images and video sequences.

At the core of the work is the application of statistical and probabalistic techniques. The group seeks to develop models that represent the types of variation exhibited in images of interest, and use these models to extract useful information from the images. We have particular strengths in constructing statistical models of shape and appearance, and in using algorithmic methodologies for design and testing of vision systems.

Summary of current active research


Dr Neil Thacker
Analysis of MR images, neural networks, stereo vision and object recognition using statistical methodologies.
Dr Thacker is a strong proponent of the need for algorithmic methodologies for design and testing.

Professor Tim Cootes

  • Development of statistical models of shape and appearance
  • Development of efficient algorithms to match such models to images and image sequences
  • Development of algorithms to construct such models automatically from unlabelled data

Professor Cootes also heads the face recognition and visual processing group, and has a long history of developing and applying statistical models of appearance to many different problems. Applications in industrial inspection, medical image analysis and face image interpretation.

Professor Chris Taylor

  • Development of statistical models of appearance and behaviour
  • Development of efficient algorithms to match such models to images and image sequences
  • Modelling and synthesizing texture
  • Applications in industrial inspection, medical image analysis and face image interpretation.

Dr Jim Graham

  • Analysis of biological and medical images.
  • Industrial inspection