The work is opening up the prospect of outdoor sporting, musical or other TV coverage that blends the excitement of being live with the spectacular visual impact that computer graphics can create. It can also be applied at the consumer level, e.g. to enable interior design ideas to be visualised by adding virtual furniture to the view of a room provided by a hand-held camera as it moves.
The system is able to work out in real-time where a camera is and how it is moving, simultaneously constructing a detailed visual map of its surroundings. This enables computer graphics to be overlaid accurately onto live pictures as soon as they are produced. Previously the blending of live action and computer-generated images has only been possible in controlled studio environments.
Harnessing techniques from mathematics, computing and engineering, the new system is being developed at Oxford University with funding from the Engineering and Physical Sciences Research Council (EPSRC). The project aims to extend the capabilities of a prototype system developed by the same team, also with EPSRC funding.
Dr Ian Reid and Dr Andrew Davison of Oxford University's Department of Engineering Science are leading the project. Andrew says: "This localisation and mapping technology turns a camera into a flexible, real-time position sensor. It has all kinds of potential applications."
The system comprises a mobile video camera connected to a laptop computer, which analyses the images it receives using software developed by the researchers. As the camera moves, the system picks out landmarks as reference points and makes a map of their 3D locations against which to measure its position. The challenge is to estimate accurately the camera's position and the layout of its surroundings at the same time - a task known as Simultaneous Localisation and Mapping (SLAM).
As well as TV and video applications, the technology under development could provide low-cost, high-performance navigation for domestic robots. It could also be incorporated into video games or wearable computing, e.g. for use in dangerous environments, where it could confirm the wearer's location and allow relevant guidance to be overlaid onto their view of surroundings.
Notes for Editors
The 3-year project, "Real-Time Camera Localisation in Real Environments", began in January 2005 and is receiving EPSRC funding of nearly £255,000. The previous 15-month study, "Real-Time Ego-Motion Estimation for a Single Camera", received EPSRC funding of nearly £61,000 and was completed in May 2004.
To deliver SLAM, the project is using an efficient probabilistic approach which can achieve real-time operation.
The project team is currently collaborating with Japan's AIST Research Institute on the development of humanoid robots.
The Engineering and Physical Sciences Research Council (EPSRC) is the UK's main agency for funding research in engineering and the physical sciences. The EPSRC invests more than £500 million a year in research and postgraduate training, to help the nation handle the next generation of technological change. The areas covered range from information technology to structural engineering, and mathematics to materials science. This research forms the basis for future economic development in the UK and improvements for everyone's health, lifestyle and culture. EPSRC also actively promotes public awareness of science and engineering. EPSRC works alongside other Research Councils with responsibility for other areas of research. The Research Councils work collectively on issues of common concern via Research Councils UK. Website address for more information on EPSRC: www.epsrc.ac.uk/
For more information, contact:
Dr Andrew Davison, Department of Engineering Science, University of Oxford, Tel: 01865-283-179 / 283-042 (mobile: 07813 800849), 273000, e-mail: email@example.com Dr Ian Reid Tel: 0186-528-3059, e-mail: firstname.lastname@example.org
Three images are available ('Map', 'Shelves' and 'Shelvespoints'from the EPSRC Press Office, contact Natasha. Richardson, tel:01793 444404, e-mail: email@example.com
Map.jpg: '3D map-making, the map has been built up from the camera's view of its surroundings'
Shelves.jpg and Shelvepoints.jpg: Kitchen scenes, an off-the-shelf solution - the new technology overlays virtual shelves and a virtual table on to a real kitchen.