Digital Image and Video Processing
The Digital Image and Video Processing group researches technologies for analysing and processing single images or image sequences (such as video or image slices) to identify, segment and manipulate objects within them. It uses a range of Digital Signal Processing (DSP) techniques on image applications and works closely with the Acoustics and Digital Music Processing research group, which employs similar algorithms and techniques in its own work.
The Digital Image and Video Processing group also addresses issues arising from the distribution of video over networks.
Research staff include: Cham Athwal, Ian Williams and Jerry Foss
Current projects
- Segmentation of teeth structures from CT and MRI scans – work being carried out jointly with Birmingham City University’s Faculty of Health and with the King’s College, London School of Dentistry
- Evaluation of Edge Detection methods
- The development of statistical edge detection algorithms for complex 2D and 3D data sets, and a further project assessing the suitability of image processing for object replacement in both image and video frames
- Machine vision project with Hadley Industries to monitor profiles of large cold-rolled steel sections
- Personalised placement of digital objects in video scenes on a commercial agent-based brokerage platform – collaboration with Instituto Superior de Engenharia do Porto (ISEP), Portugal
Recent projects
- Video over networks – KTP Industrial project with Gas St Works Ltd
- Tracking of objects in video for automotive applications
- Synchronisation of multiple media streams, including video and discrete data
- Automated segmentation and shape classification systems for analysis and object recognition, and statistical edge detection for a variety of low-level image processing applications
Resources
Statistical edge detection particularly suited for textured images: The algorithms presented in these files are the product of the thesis titled “Edge Detection of Textured Images Using Multiple Scales and Statistics”. They apply several two-sample tests to 2D image data to locate edge information.
Multiple Scale and Multiple Statistical Edge Detectors: The algorithms presented in these files extend the 2D edge detectors with the use of Artificial Neural Networks (ANN). The ANN application allows both the training and classification of edges of varying scales within an image, using a variety of statistical tests. These algorithms can be trained around a specific application and therefore tailored to a given type of data.
The Greyscale Figure of Merit (GFOM): A novel metric for assessing both the location accuracy and connectivity of edge detected images.
For more information on these resources, contact Dr Ian Williams
Recent publications
I. Williams, D. Svoboda and N. Bowring, “A Novel Performance Metric for Grey-Scale Edge Detection” – International Conference on Computer Vision Theory and Applications 2010
J. D. Foss, B. Malheiro, “Media Component Brokerage”; Proceedings of the Fourth European Conference on the Use of Modern Information and Communication Technologies – ECUMICT 2010, Gent, Belgium, March 2010
Elson, B., Athwal C., Reynolds P., “Creating the World of Augmented Dental Training”, E-Learn 2009, World Conference on E-Learning in Corporate, Government and Healthcare, AACE Vancouver, 2009
I. Williams, “Edge Detection of Textured Images Using Multiple Scales and Statistics”. PhD thesis, 2008
N. Bowring, I. Williams, C. Johnson and J. Jaiswal, “Fatigue Crack, Squat and Wheel Burn Detection by a Multi-Scale Statistical Image Processing Technique”; Proceedings of the 33rd Annual General Meeting of the British Institute of Non-Destructive Testing, 2008
I. Williams, D. Svoboda, N. Bowring and E. Guest, “Statistical Edge Detection of Concealed Weapons Using Artificial Neural Networks”. In Proceedings of SPIE-IS&T Electronic Imaging. Vol. 6812. Bellingham, Washington: SPIE, 2008; p68121J-1-12, 12 pp. ISSN 0277-786X
I. Williams, D. Svoboda, N. Bowring and E. Guest, “Improved Statistical Edge Detection Through Neural Networks”. In 10th Conference on Medical Image Understanding and Analysis 2006. ISBN: 1-901727-31-9; p56-60
D. Svoboda, I. Williams, N. Bowring and E. Guest, “Statistical Techniques for Edge Detection in Histological Images”. In 1st Int. Conf. on VISAPP – International Conference on Computer Vision Theory and Applications 2006. ISBN: 972-8865-40-6; p457 462
I. Williams, N. J. Bowring, E. Guest, P. Twigg, Y. Fan and D. Gadsby, “A Combined Statistical/Neural Network Multi-Scale Edge Detector”. In 5th IASTED Visualization Imaging and Image Processing Conf. 2005. ISBN: 0-88986-528-0; p480-266
D. Gadsby, P. Twigg, N. Bowring and I. Williams, “Ultra Wideband Positioning for Intelligent Security Systems”. The Journal for the Institute of Measurement and Control. Volume 38, p140-146. 2005.
Robinson J.E., Athwal C.S. and van Reeven V., “Analysing Engine Behaviour Through High Speed Video”, Engine Expo, Stuttgart, 2005
J. D. Foss, “Dynamic Intelligent Intermediation”; invited presentation to BBC Technology Forum, December 2005
J. D. Foss, “Beyond MultiPlay”; invited presentation to BBC Technology Forum, February 2005
J. D. Foss, “From Triple Play to the Global Jukebox”; invited presentation to BBC Technology Forum, July 2004
Robinson J.E. and Athwal C.S., “Measurement of Accessory Drivebelt Oscillations via High Speed Imaging”, Advanced Powertrain Control Symposium, Birmingham, 2004
Athwal C.S. and Robinson J., “Synchronised Multimedia for Engineering and Scientific Analysis”, Multimedia Systems Journal, 9, 365-377