| EPSRC Reference: |
GR/T21790/01 |
| Title: |
Markerless Motion Capture for Humans in Video |
| Principal Investigator: |
Torr, Professor PH |
| Other Investigators: |
|
| Researcher Co-Investigators: |
|
| Project Partners: |
|
| Department: |
Faculty of Technology, Design and Enviro |
| Organisation: |
Oxford Brookes University |
| Scheme: |
First Grant Scheme Pre-FEC |
| Starts: |
21 October 2004 |
Ends: |
20 October 2007 |
Value (£): |
124,111
|
| EPSRC Research Topic Classifications: |
|
| EPSRC Industrial Sector Classifications: |
| Creative Industries |
Information Technologies |
|
| Related Grants: |
|
| Panel History: |
|
|
Summary on Grant Application Form |
|
This proposal concerns the analysis of humans in videos. This is motivated by many possible applications for instance, motion capture, medical biomechanical analysis, ergonomic design, human computer interface, surveillence and automated pedestrian detection, sports analysis, broadcast TV analysis etc.The first task is to provide a system that will robustly detects humans in cluttered scenes, which is still an unsolved problem. The input to this system is a sequence of images provided from a normal video camera. The output is to determine for each image whether it contains a human and if so how many and at what position. The research will focus on developing new methods to do this based on recent research in machine learning, computer vision and Bayesian methods.The next task is to determine the pose of the human within the image, the pose can be specified by the angles between the joints e.g. the angles at the elbows and knees etc. This process is often referred to as 'motion capture' in which some avatar or computer graphics (CG) model is animated (put into the same pose as) a human. It is commonly used in the film industry to bring to life CG models but has so far only been achievable by attaching distinctly coloured markers to the subject which can be tracked and their motion used to infer the pose; generally this is done in a 'blue screened' studio (i.e. a special environment with carefully controlled colour making up the walls). What is aimed at here, by contrast is a markerless system that does not require a special environment, and so can be used in any location with the minimum of preparation. Indeed this technology could also be applied to any already existing footage of a human subject (or non-human animal, the techniques developed will be general).
|
|
Key Findings |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Potential use in non-academic contexts |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Impacts |
|
| Description |
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk |
| Summary |
|
| Date Materialised |
|
|
|
|
Sectors submitted by the Researcher |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
| Project URL: |
|
| Further Information: |
|
| Organisation Website: |
http://www.brookes.ac.uk |