| EPSRC Reference: |
GR/T11883/01 |
| Title: |
Novel scalable multi-view representations for compression and view synthesis |
| Principal Investigator: |
Canagarajah, Professor N |
| Other Investigators: |
|
| Researcher Co-Investigators: |
|
| Project Partners: |
|
| Department: |
Electrical and Electronic Engineering |
| Organisation: |
University of Bristol |
| Scheme: |
Standard Research (Pre-FEC) |
| Starts: |
01 June 2005 |
Ends: |
31 May 2008 |
Value (£): |
159,955
|
| EPSRC Research Topic Classifications: |
| Image & Vision Computing |
Multimedia |
|
| EPSRC Industrial Sector Classifications: |
|
| Related Grants: |
|
| Panel History: |
|
|
Summary on Grant Application Form |
|
Future advances in high definition television (HDTV), display technology and compression technologies will not only enable a higher quality of service of video, but will enable new immersive multimedia services to be supported. These applications usually require multi-view video sequences displayed in large screens to provide increased realism and 3D modelling capabilities. However, one of the key challenges for all multimedia network service providers is efficient and effective multi-view video distribution across heterogeneous networks with varying transmission capacity and quality. In this context, this proposal addresses innovative 3D representation and compression schemes for future visual communication systems, and allied algorithms. We shall investigate efficient multi-view video representations for transmission purposes while supporting high-quality view synthesis (henceforth VS) from arbitrary viewpoints.Significant advances have been achieved in developing embedded image and video coding schemes based on wavelets, producing excellent rate-distortion performance and scalable functionality with unpredictable channel variations. However, a similar approach for multi-view image sequences does not exist. We shall develop new approaches for embedded coding of multi-view sequences based on multi-dimensional spatio-temporal volumes.Embedded video compression schemes are currently designed to optimise visual quality and do not model 3D scenes making them suboptimal for arbitrary view synthesis (rendering images from different viewpoints without building explicit 3-D models). Similarly, decomposing an image sequence into depth layers, which has several advantages for scalable video transmission, has been proposed as a representation for image-based rendering (IBR) in vision and graphics, but coding issues have not been integrated so far. We shall investigate and assess comparatively the benefits of layered and embedded representations for multi-view analysis and synthesis in visual communication scenarios, thus integrating IBR and communication aspects for the first time.
|
|
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.bris.ac.uk |