BMVA/EPSRC Summer School in Computer Vision and Digital Image Computing 2010

This proposal concerns funding for a training Summer School in Computer Vision and Digital Image Computing. EPSRC has previously supported a number of Summer Schools in this area. Here we set out the case for further training support. In summary, we show that the strategic case for advanced training in this area is increasing: digital images and the need to transmit, process and interpret them are becoming an increasingly important economically and socially. We further argue that doctoral training programmes in this field cannot provide research students with a sufficient breadth of knowledge to equip them fully to contribute to the technological developments that can be anticipated.Digital images are everywhere. Increasingly images play a central role of our social and economic lives. Obvious examples are security video, biometrics, vehicle recognition, videophone calls using webcams, entertainment and medical imaging. The capacity of images to encode and transmit huge quantities of useful information means that the capture, processing and interpretation of images will form a significant component in the digital economy.

Digital images are everywhere. Increasingly images play a central role of our social and economic lives. Obvious examples are security video, biometrics, vehicle recognition, videophone calls using webcams, entertainment and medical imaging. The capacity of images to encode and transmit huge quantities of useful information means that the capture, processing and interpretation of images will form a significant component in strategic areas such as the digital economy and healthcare. The Digital Economy Already image processing and interpretation are key elements in the economically important creative industries. UK companies such as 2d3 Ltd (Oxford) and Image Metrics (Manchester) provide world leading innovative solutions in the film and games sector. In the surveillance field, a computer vision based system developed by Ipsotek is currently used by Transport for London for automatic road traffic monitoring. A number of companies develop and install systems for industrial inspection and robotics in manufacturing, These are examples of the economic impact of digital image interpretation the current economy. The importance of these areas would have been hard to predict a decade ago. Likewise it is difficult to predict what imaging applications will be significant in the near future: examples might include biometrics for personal security on mobile devices or photorealistic personal avatars in computer games. In addition to those companies involved in developing technologies for digital image manipulation, there will be a further economic impact arising from users of the technology such as manufacturers using automated inspection systems and pharmaceutical companies using digital imaging for research and development. The economic impact of digital images extends to suppliers of ancillary equipment and software, such as lighting, sensors, microscope automation, OCR, OEM hardware and software, etc. In all of these areas there is a requirement for highly trained staff with a broad understanding of the mathematical and computational methods required in processing and interpreting images. The achievements of the companies mentioned above would not be possible without a strong UK research base in Computer Vision and Digital Imaging, as they are all employers of PhD graduates in this area, some are spin-offs from major Computer Vision research groups at UK universities, or engaged in collaborations with such groups. Healthcare The expansion of the use of digital imaging in healthcare matches that in the creative and communication industries. The increasing ability to image anatomy and physiology results in an ever greater requirement for methods of interpretation, quantification and visualisation. Computer Aided Diagnostic systems are becoming common in a number of clinical disciplines. The move towards digital health records is driven by the need for efficient communication between medical and healthcare professionals, but will require a commensurate increase in the capacity for transmission, display and manipulation of larger (3D, 4D) medical images. As in the creative sector, UK companies of a range of sizes are currently addressing these issues. As in the creative sector, they include spin-offs from and collaborators of Medical Image Analysis groups in UK universities and require highly trained PhD graduates.