EP0469648A2
Expired / Lapsed
EP
⏱️ 0.0 years remaining
Data compression using a feedforward quantization estimator
Assignee: Ampex Corp
•
Filing: 1991-06-10
•
Granted: —
•
Expires: 2011-06-10
Abstract
Abstract An image data compression technique utilizes data forward feeding to estimate one or more quantization parameters for such data. The estimation process preferably is iterated a number of times, the value found at each estimation being used as the trial value for subsequent estimations. The initial trial value is selected by a data feed forward technique which assures that its value is within range of the final quantization parameter that is desired.
Inventors
- 👤 Sidney D. Miller
- 👤 Peter Smidth
- 👤 Charles H. Coleman
Bibliographic Details
Application Number
—
Jurisdiction
EP
Data Source
google_patents
Last Synchronized
2026-10-29 04:36
CPC Classifications
H; H03; H03M; H03M7/00; H04; H04N; H04N19/00; H04N19/85; H04N19/88; H04N19/10; H04N19/102; H04N19/124; H04N19/126; H04N19/134; H04N19/146; H04N19/149; H04N19/169; H04N19/17; H04N19/172; H04N19/176; H04N19/186; H04N19/189; H04N19/192; H04N19/42; H04N19/46; H04N19/60
Claims Analysis
Total Claims: 22
Claim 1
Independent Claim
161 words
A method of preparing a digital video signal, having a data rate (Din) for transmission as an encoded output signal having a data rate (D out ) which is less than the input data rate (Din) said method comprising the steps of: transforming said video signal to a spatial-frequency domain digital signal formatted in groups of transform coefficients; quantizing said predetermined groups of transform coefficients to integral values in accordance with a quantizing parameter (Q f ) to generate quantized coefficients; encoding said quantized coefficients to produce said output signal having the output data rate (D out ); and selecting, prior to the quantizing step, the quantizing parameter (Q f ) for each group of transform coefficients, said quantizing parameter (Q f ) being selected to generate an encoded output signal having a number of encoded bits corresponding to the desired output data rate (D out ), and feeding each selected quantizing parameter (Q f ) forward to said quantizing step.
Claim 2
Independent Claim
161 words
A method of preparing a digital video signal, having a data rate (Din) for transmission as an encoded output signal having a data rate (D out ) which is less than the input data rate (Din) said method comprising the steps of: transforming said video signal to a spatial-frequency domain digital signal formatted in groups of transform coefficients; quantizing said predetermined groups of transform coefficients to integral values in accordance with a quantizing parameter (Q f ) to generate quantized coefficients; encoding said quantized coefficients to produce said output signal having the output data rate (D out ); and selecting, prior to the quantizing step, the quantizing parameter (Q f ) for each group of transform coefficients, said quantizing parameter (Q f ) being selected to generate an encoded output signal having a number of encoded bits corresponding to the desired output data rate (D out ), and feeding each selected quantizing parameter (Q f ) forward to said quantizing step.
Claim 3
Dependent Claim
165 words
A method according to claim 1 wherein said selecting step further comprises a series of steps, namely: delaying, subsequent to said transforming step, said transform coefficients for a first predetermined period of time; selecting, while delaying said transform coefficients, a trial quantizing parameter and counting the number of encoded bits that would be produced by said encoding step during said first predetermined period of time if (Q t ) the trial quantizing parameter were used as the quantizing parameter (Q f ); determining the difference between said counted number of encoded bits and a number of bits corresponding to the output data rate (D out ); applying, if said difference is zero, said trial quantizing parameter (Q t ) in said quantizing step for use as said quantizing parameter (Q f ); and, computing, if said difference is greater than zero, an updated value for said trial quantizing parameter (Q t ) so as to reduce said difference, and repeating the said series of steps.
Claim 4
Dependent Claim
21 words
A method according to claim 2 wherein said transforming step further comprises transforming said video signal using a discrete cosine transform.
Claim 5
Dependent Claim
128 words
A method according to claim 3 wherein the digital video signal comprises a succession of fields, the method further comprising, at the beginning of each field of said digital video signal, the steps of: delaying, prior to the said first delaying step, said groups of transform coefficients for a time corresponding to a multiple (N) of the first predetermined period of time, while also performing the said series of steps N times on the same set of transform coefficients; and applying the trial quantizing parameter (Q t ) selected on the Nth selecting step in the said delaying step as the trial quantizing parameter (Q t ) for the first selecting step for all subsequent trial quantizing parameters related to a particular field of said digital video signal.
Claim 6
Independent Claim
31 words
A method according to any foregoing claim wherein said encoding step further comprises encoding said quantized coefficients using variable length coding to produce an output signal having the desired data rate.
Claim 7
Independent Claim
17 words
A method according to any foregoing claim wherein each of said steps is performed in real time.
Claim 8
Independent Claim
58 words
A method according to any foregoing claim further comprising the steps of: formatting subsequent to said encoding step said encoded output signal into the format of said channel; and interleaving with said formatted encoded output signals said quantizing parameter (Q f ) with each encoded group of quantized coefficients to which said quantizing parameter (Q f ) relates.
Claim 9
Dependent Claim
70 words
Apparatus for compressing a data signal defining a video image, comprising: means (15) for transforming said data signal to a spatial-frequency domain signal formatted in predetermined groups of transform coefficients; means (9,40) for using at least one of said predetermined groups of transform coefficients to estimate the value of a numerical processing parameter for quantizing said transform coefficients; and means (40,45) for applying said estimate subsequently to said transform coefficients.
Claim 10
Dependent Claim
70 words
Apparatus for compressing a data signal defining a video image, comprising: means (15) for transforming said data signal to a spatial-frequency domain signal formatted in predetermined groups of transform coefficients; means (9,40) for using at least one of said predetermined groups of transform coefficients to estimate the value of a numerical processing parameter for quantizing said transform coefficients; and means (40,45) for applying said estimate subsequently to said transform coefficients.
Claim 11
Dependent Claim
32 words
Apparatus according to claim 8 wherein said means (40,45) for applying said estimate is adapted to apply the same to the very same portion of said data used to determine said estimate.
Claim 12
Dependent Claim
22 words
Apparatus according to claim 9 wherein said means (40,45) for applying said estimate also is adapted to do so in real time.
Claim 13
Independent Claim
163 words
A method of preparing a digital signal, having a data rate (Din) for transmission as an encoded output signal, having an output data rate (D out ) which is less than the input data rate (Din), said method comprising the steps of: transforming said digital signal to a spatial-frequency domain digital signal formatted in predetermined groups of transform coefficients; quantizing said predetermined groups of transform coefficients to integral values in accordance with a quantizing parameter (Q f ) to generate quantized coefficients; encoding said quantized coefficients to produce said output signal having the output data rate (D out ); and selecting, prior to the quantizing step, the quantizing parameter (Q f ) for each predetermined group of transform coefficients, said quantizing parameter (Q f ) being selected to generate an encoded output signal having a number of encoded bits corresponding to the desired output data rate (D out ); and feeding each selected quantizing parameter (Q f ) forward to said quantizing step.
Claim 14
Independent Claim
163 words
A method of preparing a digital signal, having a data rate (Din) for transmission as an encoded output signal, having an output data rate (D out ) which is less than the input data rate (Din), said method comprising the steps of: transforming said digital signal to a spatial-frequency domain digital signal formatted in predetermined groups of transform coefficients; quantizing said predetermined groups of transform coefficients to integral values in accordance with a quantizing parameter (Q f ) to generate quantized coefficients; encoding said quantized coefficients to produce said output signal having the output data rate (D out ); and selecting, prior to the quantizing step, the quantizing parameter (Q f ) for each predetermined group of transform coefficients, said quantizing parameter (Q f ) being selected to generate an encoded output signal having a number of encoded bits corresponding to the desired output data rate (D out ); and feeding each selected quantizing parameter (Q f ) forward to said quantizing step.
Claim 15
Dependent Claim
67 words
A video data compression system wherein serial data defines sequential active and inactive display occurrences, characterised by: means (20) for delaying said data; means for sampling data defining an active occurrence; and means (25,40) using said sampled data during the delay of data defining an inactive occurrence to calculate an initial trial value for a numerical processing parameter to be applied to data defining said active occurrence.
Claim 16
Dependent Claim
67 words
A video data compression system wherein serial data defines sequential active and inactive display occurrences, characterised by: means (20) for delaying said data; means for sampling data defining an active occurrence; and means (25,40) using said sampled data during the delay of data defining an inactive occurrence to calculate an initial trial value for a numerical processing parameter to be applied to data defining said active occurrence.
Claim 17
Dependent Claim
26 words
A video data compression system according to claim 12 wherein said means for sampling samples data defining an active occurrence which immediately follows said inactive occurrence.
Claim 18
Dependent Claim
82 words
A compression system for image data defining sequential active and inactive image display occurrences, characterised by: means for passing said data sequentially through a channel; means for feeding forward a portion of said data defining an active display occurrence during the time a portion of said data defining an inactive display occurrence is in said channel; and means for using said data portion to estimate the value of a numerical processing parameter to be applied to data defining an active display occurrence.
Claim 19
Dependent Claim
82 words
A compression system for image data defining sequential active and inactive image display occurrences, characterised by: means for passing said data sequentially through a channel; means for feeding forward a portion of said data defining an active display occurrence during the time a portion of said data defining an inactive display occurrence is in said channel; and means for using said data portion to estimate the value of a numerical processing parameter to be applied to data defining an active display occurrence.
Claim 20
Dependent Claim
59 words
A compression system according to claim 14 wherein said means for feeding forward a portion of said data includes means for delaying said data during passage through said channel of that portion of the same defining an inactive display occurrence, and means for sampling data defining the active display occurrence immediately following said inactive display occurrence during said time.
Claim 21
Independent Claim
53 words
A method of processing data, characterised by: using said data to calculate a plurality of successive estimated values (Q t ) of a numerical processing parameter to be applied to said data, each of said values being estimated with the value calculated from an earlier estimation being used as an initial trial value.
Claim 22
Independent Claim
53 words
A method of processing data, characterised by: using said data to calculate a plurality of successive estimated values (Q t ) of a numerical processing parameter to be applied to said data, each of said values being estimated with the value calculated from an earlier estimation being used as an initial trial value.
References Cited (Backward Citations)
No citation records indexed.
USPTO Annuity & Maintenance Windows
Maintenance schedule is calculated for granted US utility patents.