Device for generating multiple quality level bit-rates in a video encoder

• US 6,501,798 B1
• Filed: 05/15/2000
• Issued: 12/31/2002
• Est. Priority Date: 01/22/1998
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Sivan, Zohar
• Primary Examiner(s): BRITTON, HOWARD W
• Application Number:
  US09/570,806
• Time in Patent Office: 960 Days
• Field of Search: H04/N.750, 375/240.12, 375/240.14, 375/240.04, 358/426, 382/162, 382/166, 382/232, 382/237, 382/238, 382/240, 382/251
• US Class Current: 375/240.12
• CPC Class Codes: H04N 19/126 Details of normalisation or weighting functions, e.g. normalisation matrices or variable uniform quantisers
• CPC Class Codes: H04N 19/137 Motion inside a coding unit, e.g. average field, frame or block difference
• CPC Class Codes: H04N 19/149 by estimating the code amount by means of a model, e.g. mathematical model or statistical model
• CPC Class Codes: H04N 19/172 the region being a picture, frame or field
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/177 the unit being a group of pictures [GOP]
• CPC Class Codes: H04N 19/187 the unit being a scalable video layer
• CPC Class Codes: H04N 19/36 Scalability techniques involving formatting the layers as a function of picture distortion after decoding, e.g. signal-to-noise [SNR] scalability
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

Device for generating multiple quality level bit-rates in a video encoder having a motion estimator providing a predicted block for each predefined block based upon estimating the motion between the predefined block of the current image and the corresponding block in the previous image, a transformer for transforming a prediction error resulting from the difference between the predicted block and the predefined block into the frequency domain, and a quantizer for quantizing the coefficients of the prediction error and providing the quantized coefficients to a video multiplex coding unit. Such a device includes a number n of stages, each corresponding to a quality level i=1 to n, and each stage having a computer for reducing the prediction error in accordance with the quality level i and providing a corresponding quantized prediction error residual QDi, and an adder for obtaining a cumulative prediction error: ${\mathrm{QD}}_{\mathrm{TOTAL}}=\sum _{j=0}^{j=i}{\mathrm{QD}}_j$

corresponding to quality level i

wherein QD_j is the de-quantized value of the quantized prediction error residual QD_j. Such a device is well suited to the current H.263 bitstream structure of the ITU recommendation H.324/H.323 by using the sub-bitstream feature without the need to change the bitstream structure.

Claims

1. In a video encoder comprising motions estimation means (22) providing a predicted block for each predefined block based upon estimating the motion between said predefined block of a current image and the corresponding block in a previous image, transform means (16) for transforming a prediction error resulting from the difference between said predicted block and said predefined block into the frequency domain, and quantizing means (20) for quantizing coefficients of the prediction error in the frequency domain and providing the quantized coefficients to a video multiplex coding unit (30), wherein said quantized coefficients are de-quantized (24) and inverse transformed (26) to give back said prediction error and add it to said predicted block whereby the result is provided to said motion estimation means in order to get a new current predicted block;

a device comprising means for generating from one video sequence, multiple video bitstreams of different bit-rates and corresponding to different quality levels, said device further comprising;

means for building a hierarchical H.263 bitstream including at least one and up to four substreams each corresponding to one different quality level generated from said video sequence, a CPM bit of the H.263 bitstream being set to 1, a PSBI 2 bit field being set to the number of different quality level substreams stored, a first Group Of Block comprising a 2 bit field GSBI being set to the first quality level bitstream, the corresponding Macroblock layer comprising MVD, MVD2, MVD3, MVD4, MVDB bit fields storing motion information generated from said video sequence for said first quality level bitstream, following Group Of Blocks comprising a 2 bit field GSBI being set to the value corresponding to the range of the following quality level bitstreams wherein the corresponding Macroblock layer MVD, MVD2, MVD3, MVD4, MVDB bit fields do not store motion information.

2. Device according to claim 1 further comprising:

a number n of stages, each said stage corresponding to a quality level i=1 to n, each said stage comprising computing means for reducing the prediction error in accordance with said quality level i and providing a corresponding quantized prediction error residual QD_i, and

summing means for getting a cumulative prediction error ${\mathrm{QD}}_{\mathrm{TOTAL}}=\underset{j=0}{\overset{j=i}{\sum <br><br>}}<br><br>{\mathrm{QD}}_j$

corresponding to quality level i.

wherein QD_j is a dequantized value of the quantized prediction error residual QD_j.

3. Device according to claim 2, wherein said computing means in each one of said stages comprise:

means for determining a prediction error residual E_i corresponding to quality level i, said residual being the difference between prediction error residual E<sub>i−

1</sub> of quality level i−

1 and the de-quantized value of the quantized prediction error residual QD<sub>i−

1</sub> obtained in the stage corresponding to quality level i−

1, and

quantizing means for quantizing said prediction error residualE_i according to a quantization step size smaller than the quantization step size of the stage corresponding to quality level i−

1.

4. Device according to claim 3, wherein said quantizing means in each stage corresponding to quality level i is a quantizer (40, 46, 52 or 58) having a quantization step size Q_i smaller than the quantization step size Q_i−

1 of the stage corresponding to quality level i−

1.

5. Device according to claim 4, wherein in each stage corresponding to quality level i, the quantized prediction error residual QD_i is de-quantized in an inverse quantizer (42, 48, or 54) having a quantization step size IQ_i equal to the quantizer step size Q_i used in quantizer of said stage.

6. Device according to any one of claims 5, wherein the input of the first stage corresponding to quality level i=1 is said prediction error and the output of said stage being prediction error residual E_i.

7. Device according to claim 6, wherein the de-quantized values QD_i for i=0 to 3 are respectively obtained by means of inverse quantizers using the same quantization step sizes IQ₀, IQ₁, IQ₂ or IQ₃ as the quantization step sizes Q₀ to Q₃ of quantizers providing said quantized prediction error QD₀ and quantized prediction error residuals QD₁, QD₂ or QD₃.

8. In a video encoder comprising motions estimation means (22) providing a predicted block for each predefined block based upon estimating the motion between said predefined block of the current image and the corresponding block in the previous image, transform means (16) for transforming a prediction error resulting from the difference between said predicted block and said predefined block into the frequency domain, and quantizing means (20) for quantizing the coefficients of the prediction error in the frequency domain and providing the quantized coefficients to a video multiplex coding unit (30), wherein said quantized coefficients are de-quantized (24) and inverse transformed (26) to give back said prediction error and add it to said predicted block whereby the result is provided to said motion estimation means in order to get a new current predicted block;

a device comprising means for generating a video sequence from reading a bitstream of a compressed video sequence of different specific bit-rate and corresponding to a specific quality level, said device further comprising;

means for reading a hierarchical H.263 bitstream including at least one and up to four substreams each corresponding to one different bit-rate corresponding to different quality level generated from one video sequence, a CPM bit of the H.263 bitstream being set to 1, a PSBI 2 bit field being set to the number of different quality level substreams stored,the first Group Of Block comprising a 2 bit field GSBI being set to X‘

01’

for the first quality level bitstream, the corresponding Macroblock layer comprising MVD, MVD2, MVD3, MVD4, MVDB bit fields storing motion information for said first subitstream, and the following Group Of Blocks if any, comprising a 2 bit field GSBI being set to the value corresponding to the range of the corresponding quality level bitstream, the motion information being the content of the MVD, MVD2, MVD3, MVD4, MVDB bit fields read in the first subitstream.