Details of Grant 

EPSRC Reference:	GR/J08386/01
Title:	CODING OF SUPER HIGH DEFINITION (SHD) IMAGES
Principal Investigator:	Pearson, Professor D
Other Investigators:	Ghanbari, Professor M Clark, Dr AF
Researcher Co-Investigators:
Project Partners:
Department:	Computing and Electronic Systems1
Organisation:	University of Essex
Scheme:	Standard Research (Pre-FEC)
Starts:	01 September 1993	Ends:	31 January 1997	Value (£):	308,069
EPSRC Research Topic Classifications:	Digital Signal Processing
EPSRC Industrial Sector Classifications:
Related Grants:
Panel History:
Summary on Grant Application Form
To develop efficient, transparent coding methods for still and moving Super High Definition (SHD) images, in the context of appropriate areas of application.Progress:The initial stages of the project involved setting up the super-high-definition (SHD) equipment required for the work, including an SHD monitor, a specially-constructed frame store, and controlling computer. Necessary software was also developed to establish a fully-functioning environment for SHD image coding research. Coding of still SHD images was investigated first. Hierarchical coding algorithms are thought to be most suitable for SHD images since they provide the capability for progressive transmission, and compatibility with lower-resolution images and multimedia data. Two pyramidal algorithms, the Laplacian and discrete cosine transform (DCT) pyramids, have been investigated in detail to date. The Laplacian pyramid is a decomposition that takes place in the spatial domain, where more and more detail is added to a low-pass filtered, low-resolution image to recover the original SHD image. Conversely, this pyramidal decomposition is carried out in the transform domain in the DCT pyramid algorithm. Work concentrated on optimising quantisation of the different levels of the pyramids, which is the key to obtaining high compression, whilst simultaneously achieving a reconstructed image that is subjectively loss-less (often termed perceptually transparent ). Both pyramidal algorithms compress images effectively, with the DCT pyramid giving slightly better performance at the expense of more computation. Good quality reconstruction has been achieved at half a bit per pixel, and perceptually transparent coding at one bit per pixel. A paper on this work was presented at the 1994 International Picture Coding Symposium in Portland, USA, and another has been submitted to the 1995 EE Conference on Image Processing and its Applications. Multimedia images represent a major potential application area for SHD images. A multimedia image often contains text, drawings, graphs, etc., as well as natural scenes. Current coding schemes, like the DCT-based JPEG method, code natural scenes efficiently but perform less well on graphics. A graphics-segmentation based coding scheme has been developed to code multimedia images. It first segments an image into graphics and natural scene areas and then codes graphics with run-length coding and natural scene areas with the DCT. Sequences of SHD images have been obtained very recently from NTT in Japan, so our attention is shifting from still images to motion sequences. Work on motion compensation is currently being carried out, in particular on the separation of foreground motion and background panning, which could yield large compression. Motion compensation in the wavelet domain is also being investigated.
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.sx.ac.uk