US5351095A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method and device for estimating and hierarchically coding the motion of sequences of images
Assignee: Technicolor USA Inc
•
Filing: 1990-08-24
•
Granted: 1994-09-27
•
Expires: 2010-08-24
Abstract
Abstract A method for estimating motion which calculates a priori for a current image a first field of motion for blocks of a first level resulting from a partition of the image into blocks of a so-called maximum size, on the basis of a symmetric estimation algorithm, using the preceding and following images in the sequence to calculate the displaced inter-raster differences. Then, a second field of motion for blocks resulting from the partition of a block of the first level are calculated, and so on until a level of blocks of a minimum size. The final field of motion results from the selection of the most suited motion vectors from those of the various levels and is directly associated with a quadtree coding tree. Such a method finds particular application to image transmission chains, and in particular high-definition.
Inventors
- 👤 Michel Kerdranvat
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-08 13:41
CPC Classifications
G; G06; G06T; G06T9/00; G06T9/40; G06T7/00; G06T7/20; G06T7/223; G06T7/238; H; H04; H04N; H04N19/00; H04N19/50; H04N19/503; H04N19/51; H04N19/53; H04N19/90; H04N19/96
Claims Analysis
Total Claims: 12
Claim 1
Independent Claim
254 words
A method for estimating and hierarchically coding the motion of sequences of images, comprising the steps of: at a first step, partitioning a current image into 2 P+1 ×2 P+1 pixel macroblocks, where p is an integer, and determining first field of motion vectors, which is associated with this partition into macroblocks, by using an iterative and recursive estimation algorithm initialized by field of motion vectors estimated for a preceding image; at a second step, partitioning into quadrants each of the macroblocks and determining, for the blocks which result therefrom, second field of motion vectors by using the interactive and recursive estimation algorithm but initialized by vectors from the field of motion vectors which were estimated at the preceding level; at an ith step, i=3 to D, where n is an integer, partitioning each of the blocks of the i-1 level into quadrants and determining, for the blocks which result therefrom, an ith field of motion vector by using the iterative and recursive estimation algorithm initialized by vectors from the field of motion vectors which were estimated at the preceding level, the blocks of minimum size being 2 l+1 ×2 l+1 pixel blocks, where l is an integer; determining a final motion vector field from the n resulting field of motion vectors by choosing a lowest level of partition for which the field of motion vector associated with the corresponding block leads to a minimization of a criterion reflecting differences of brightness between blocks communicating in the successive images via the estimated displacement vectors.
Claim 2
Independent Claim
254 words
A method for estimating and hierarchically coding the motion of sequences of images, comprising the steps of: at a first step, partitioning a current image into 2 P+1 ×2 P+1 pixel macroblocks, where p is an integer, and determining first field of motion vectors, which is associated with this partition into macroblocks, by using an iterative and recursive estimation algorithm initialized by field of motion vectors estimated for a preceding image; at a second step, partitioning into quadrants each of the macroblocks and determining, for the blocks which result therefrom, second field of motion vectors by using the interactive and recursive estimation algorithm but initialized by vectors from the field of motion vectors which were estimated at the preceding level; at an ith step, i=3 to D, where n is an integer, partitioning each of the blocks of the i-1 level into quadrants and determining, for the blocks which result therefrom, an ith field of motion vector by using the iterative and recursive estimation algorithm initialized by vectors from the field of motion vectors which were estimated at the preceding level, the blocks of minimum size being 2 l+1 ×2 l+1 pixel blocks, where l is an integer; determining a final motion vector field from the n resulting field of motion vectors by choosing a lowest level of partition for which the field of motion vector associated with the corresponding block leads to a minimization of a criterion reflecting differences of brightness between blocks communicating in the successive images via the estimated displacement vectors.
Claim 3
Dependent Claim
33 words
The method according to claim 1, wherein the iterative calculations applied at a given level are performed in parallel, the calculations of a given level taking their initialization values at a preceding level.
Claim 4
Dependent Claim
42 words
The method according to claim 1, wherein p is equal to 5 and l equal to 2, the macroblocks being 128×128 pixel blocks and the blocks of minimum size being 8×8 pixel blocks, the estimation method being hierarchically structured into 5 levels.
Claim 5
Dependent Claim
57 words
The method according to claim 1, wherein a point field of motion is established from a final field of motion vector by allocating to each pixel the motion vector of the final field for which an interraster brightness difference between this pixel and the pixel displaced by the corresponding motion vector in the preceding raster is minimized.
Claim 6
Dependent Claim
83 words
The method according to claim 1, wherein the iterative and recursive estimation algorithm used is symmetric and uses, for the determination of a field of motion vector at the iteration rank i, the motion vector determined at the iteration i-1 less sums of differences of brightness between pixels of the current block and pixels of blocks corresponding to the current block which are displaced by the field of motion vector, in the preceding raster and in the following raster of the sequence respectively.
Claim 7
Independent Claim
134 words
The method according to either one of claims 1 or 5, wherein for each block, the iterative and recursive estimation algorithm is applied four times, initialized by four field of motion vectors respectively associated at the first step, with four blocks of lower level constituting each macroblock, taken from the second level field of motion vectors of the preceding image, in the following steps, with four blocks of the preceding level belonging to a neighborhood of the current block, in the same image, and in that the vector chosen from the four result vectors to form one of the vectors of the corresponding field of motion is that which leads to the minimization of the criterion reflecting the differences of brightness between blocks communicating in the successive images via the corresponding estimated displacement vectors.
Claim 8
Dependent Claim
66 words
The method according to claim 6, wherein the four motion vectors for initializing the blocks of the preceding level which are used as initial values for the estimation of a current block are the vector associated with the block of a preceding level from which it issues, and the three vectors associated with the blocks of the preceding level which are adjacent to this current block.
Claim 9
Dependent Claim
30 words
The method according to claim 6, wherein the four motion vector calculations performed for each block are performed in parallel, these calculations taking their initialization values at a preceding level.
Claim 10
Independent Claim
79 words
The method according to either of claims 1 or 5, wherein the final, hierarchical field of motion vector is directly associated with a coding tree of a quadtree type, corresponding to the hierarchy of the blocks and of the corresponding motion vectors of the various levels selected for the final field of motion sector, and in that an adjustment of the resulting coding throughput is performed by selecting the final field of motion vector suited to a required throughput.
Claim 11
Dependent Claim
40 words
The method according to claim 9, wherein the adjustment is performed, when the required throughput is exceeded, by eliminating from the final field of motion vector, least frequent vectors and by replacing them by the closest vectors of greater frequency.
Claim 12
Dependent Claim
45 words
The method according to claim 9, wherein the adjustment is performed, when the required throughput is exceeded, by removing the vectors of the field of motion which is associated with the blocks of minimum size from the selection of the final field of motion vector.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1997-09-26 | 1998-03-28 | 1998-09-27 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2001-09-26 | 2002-03-28 | 2002-09-27 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2005-09-26 | 2006-03-28 | 2006-09-27 | $7700.00 | Window Closed / Lapsed |