Patent US4894713A successfully refreshed from Google Patents.
US4894713A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method of coding video signals
Assignee: BELGIAN STATE (L'ETAT BELGE) REPRESENTED BY SECRETARY GENERAL OF SERVICES DE PROGRAMMATION de la POLITIQUE SCIENTIFIQUE
•
Filing: 1988-06-01
•
Granted: 1990-01-16
•
Expires: 2008-06-01
Abstract
Abstract In a method in which the data related to the image are read in successive blocks, then subjected to orthogonal transformation, the coefficients of the transform obtained being quantized and analyzed according to a certain order and applied in the form of binary signals to a redundancy reducing coder, the signals representing the coefficients of the transform are stored in a memory organized in several blocks, and the contents of said memory are analyzed such that the signals applied to the redundancy reducing coder comprise sets of signals of coefficients taken from a group of blocks, the signals of each coefficient in each set being taken from a separate block belonging to said group of blocks.
Inventors
- 👤 Paul Delogne
- 👤 Benoit Macq
- 👤 Luc Vandendorpe
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-24 14:07
CPC Classifications
H; H04; H04N; H04N19/00; H04N19/10; H04N19/102; H04N19/124; H04N19/126; H04N19/60; H04N19/61; H04N19/619
Claims Analysis
Total Claims: 12
Claim 1
Independent Claim
97 words
A method of coding video signals comprising: reading data relative to a a still or moving video image in successive blocks; subjecting the data to orthogonal transformation; quantizing the transformed coefficients; storing by rank signals representing the coefficients in organized blocks of memory; retrieving from memory signals representing the coefficients in groups of blocks, in a predetermined scanning order, the signals of each coefficient in each set being taken in a separate block belonging to said group of blocks; and applying the coefficients of the transformation in the form of binary signals to a variable length coder.
Claim 2
Independent Claim
97 words
A method of coding video signals comprising: reading data relative to a a still or moving video image in successive blocks; subjecting the data to orthogonal transformation; quantizing the transformed coefficients; storing by rank signals representing the coefficients in organized blocks of memory; retrieving from memory signals representing the coefficients in groups of blocks, in a predetermined scanning order, the signals of each coefficient in each set being taken in a separate block belonging to said group of blocks; and applying the coefficients of the transformation in the form of binary signals to a variable length coder.
Claim 3
Dependent Claim
89 words
Process according to claim 1, characterized in that the orthogonal transformation (T) is such that each one of the transformed coefficients (Y n .sbsb.1 n .sbsb.2) of the transform is obtained by multiplying the transformed coefficients (X n .sbsb.1 n .sbsb.2) of a unitary transform known of itself by factors (√W n .sbsb.1 n .sbsb.2) such that the quantization noise power weighted by the sensitivity of the eye as a function of spatial frequencies is minimal, this weighting being carried out independently of the statistical properties of the images.
Claim 4
Dependent Claim
40 words
Process according to claims 1 or 2, characterized in that said scanning by groups of blocks of coefficients of said perceptive transformation is carried out in such a manner as to obtain a progressively increasing quality of the coded image.
Claim 5
Dependent Claim
29 words
Process according to Claims 1 or 2, characterized in that said orthogonal transformation is applied to blocks of image elements taken in the same frame or in different frames.
Claim 6
Dependent Claim
31 words
Process according to Claims 1 or 2, characterized in that said orthogonal transformation and said scanning are applied to the predictive error determined in a known manner in a hybrid coder.
Claim 7
Dependent Claim
27 words
Process according to Claims 1 or 2, characterized in that said scanning is applied to groups of blocks comprising both image element blocks and predictive error blocks.
Claim 8
Dependent Claim
40 words
Process according to Claims 1 or 2, characterized in that said scanning presents to the redundancy reducing coder sets of bits, each bit in a set belonging to a coefficient taken in a separate block of a group of blocks.
Claim 9
Dependent Claim
21 words
Process according to Claims 1 or 2, characterized in that the perceptive transform is calculated by means of a matrix multiplier.
Claim 10
Independent Claim
120 words
A method of coding video signals comprising: reading data relative to a still or moving video image in successive blocks; subjecting the data to orthogonal transformation; quantizing the coefficients of the transformation; and applying the coefficients of the transformation to a redundancy reducing coder; wherein each one of the transformed coefficients (Y n .sbsb.1 n .sbsb.2) of the transformation is obtained by multiplying the transformed coefficients (X n .sbsb.1 n .sbsb.2) of a unitary transformation known in itself by factors (√W n .sbsb.1 n .sbsb.2) such that the quantization noise power weighted by the sensitivity of the eye as a function of the spatial frequencies is minimal, said weighting being carried out independently of the statistical properties of the images.
Claim 11
Independent Claim
120 words
A method of coding video signals comprising: reading data relative to a still or moving video image in successive blocks; subjecting the data to orthogonal transformation; quantizing the coefficients of the transformation; and applying the coefficients of the transformation to a redundancy reducing coder; wherein each one of the transformed coefficients (Y n .sbsb.1 n .sbsb.2) of the transformation is obtained by multiplying the transformed coefficients (X n .sbsb.1 n .sbsb.2) of a unitary transformation known in itself by factors (√W n .sbsb.1 n .sbsb.2) such that the quantization noise power weighted by the sensitivity of the eye as a function of the spatial frequencies is minimal, said weighting being carried out independently of the statistical properties of the images.
Claim 12
Dependent Claim
132 words
Process according to Claim 9 in which the noise weighting coefficients W n .sbsb.1 n .sbsb.2 are used to define a weighted quantization noise power given by the formula ##EQU8## where σ 2 .sub.εx, n.sbsb.1 n .sbsb.2 is the variance of the quantization error affecting the transformed coefficient X n .sbsb.1 n .sbsb.2 of the unitary transformation, characterized in that the weighting coefficients are related to a function W(u,v), expressing the sensitivity of the eye as a function of spatial frequencies u and v in cycles per height of image, by the relation ##EQU9## where r 1 and r 2 are the geometric dimensions of a block and H n .sbsb.1 n .sbsb.2 is the discrete Fourier transform of the basic order function (n 1 n 2 ) of the unitary transformation.
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) | 1993-01-15 | 1993-07-17 | 1994-01-16 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1997-01-15 | 1997-07-17 | 1998-01-16 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2001-01-15 | 2001-07-17 | 2002-01-16 | $7700.00 | Window Closed / Lapsed |