Coder/decoder for television image sub-band compatible coding, a

Television – Bandwidth reduction system – Data rate reduction

Patent

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

348413, H04N 732

Patent

active

059073608

DESCRIPTION:

BRIEF SUMMARY
BACKGROUND OF THE INVENTION

Field of the Invention
The present invention relates to a coder/decoder for sub-band compatible coding of television images and its application to hierarchical motion coding by tree structures.
It applies to digital transmission systems and especially to the transmission of digital image sequences.


DISCUSSION OF THE BACKGROUND

Recent technological progress with regard to digital transmission, together with the development of broadband digital networks, have given the role of digital compression a strategic position. Whereas the bit rates allocated for data transmission ceaselessly diminish, the requirement in relation to the quality of coding grows. Thus, in respect of the transmission of digital image sequences, the algorithms developed require the use of ever more enhanced and complex techniques, usually based on exploiting temporal redundancy together with motion estimation. This is the case for example with bidirectional coding, with "frame skip", an expression defining a technique of image jumping, or with techniques of temporal interpolation.
However, these techniques entail the appearance of new constraints as regards motion estimation. Bidirectional coding requires the ability to estimate motions of large magnitude, especially if the images processed are in the High Definition (HD) format. "Frame skip" and temporal interpolation demand accurate vector fields which are as alike the physical field as possible. Moreover, since the motion information to be transmitted becomes substantial, it is essential to reduce the cost of coding the motion field as much as possible: the most effective means of doing this is to make it as homogeneous as possible. Accuracy, homogeneity, large magnitudes and fitness to the physical field are the main characteristics to be taken into account by the motion estimator.
The majority of known image coders/decoders use discrete cosine transform coding, known by the abbreviation DCT, operating in intra and inter mode together with a motion estimation technique of "block matching" type, an expression denoting a technique according to which the image is partitioned into blocks, one motion vector per block being determined by a correlation computation, this technique making it possible to obtain motion vectors computed in terms of integer numbers of pixels, or more rarely of "pel-recursive" type, where pel is the abbreviation for pixel, developed notably in "Motion estimation and compensation for image sequence coding", published in "Image Communication 4, 1992, pp 161-174". Pel-recursive does not enable sufficient accuracy to be achieved at the level of the motion of contours and is still too sensitive to noise.
A more widespread technique which gives better results in terms of reducing the prediction error is the technique known by the expression "block matching full search". This technique does not however allow estimation of motions of large magnitude: the increase in the search window is rapidly limited by considerations of hardware implementation. Moreover, the "full search" technique does not take into account the spatial correlation between the vectors arising from adjacent blocks, this often being manifested as a noisy field which is remote from the true motion. Raising the size of the search window can only accentuate the inhomogeneity of the estimated field and, consequently, its non-negligible coding cost at low bit rate. The "full search" technique is therefore no longer suited to the new constraints.
More complex but faster "block matching" techniques permitting more substantial vector excursions have been developed in several articles and notably in a publication by Q. Wang and R. J. Clarke entitled "Motion estimation and compensation for image sequence coding" published in "Image Communication 4, 1992, pp. 161-174" and in a publication by L. Lee, J. WANG, J. Lee and J. Shie entitled "Dynamic Search Window Adjustment and Interlaced Search for Block S Hatching Algorithm" published in "International Workshop on HDTV'93, Proceedings, v

REFERENCES:
patent: 4817182 (1989-03-01), Adelson et al.
patent: 4918524 (1990-04-01), Ansari et al.
patent: 4969040 (1990-11-01), Gharavi
patent: 5097331 (1992-03-01), Chen et al.
patent: 5278915 (1994-01-01), Chupeau et al.
patent: 5420891 (1995-05-01), Akansu
patent: 5446495 (1995-08-01), Tourtier et al.
patent: 5633684 (1997-05-01), Teranishi
J. Mau et al, "Sub-band source coding for HDTV", EBU Technical Review, Spring 1992, pp. 34-44.

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Coder/decoder for television image sub-band compatible coding, a does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Coder/decoder for television image sub-band compatible coding, a, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Coder/decoder for television image sub-band compatible coding, a will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-404029

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.