Flexible quantization

• US 9,967,561 B2
• Filed: 04/28/2014
• Issued: 05/08/2018
• Est. Priority Date: 05/05/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Tu, Chengjie, Srinivasan, Sridhar
• Primary Examiner(s): PONTIUS, JAMES M
• Application Number:
  US14/263,584
• Publication Number:
  US 20140233638A1
• Time in Patent Office: 1,471 Days
• CPC Class Codes: H04N 19/124 Quantisation
• CPC Class Codes: H04N 19/126 Details of normalisation or weighting functions, e.g. normalisation matrices or variable uniform quantisers
• CPC Class Codes: H04N 19/136 Incoming video signal characteristics or properties
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/46 Embedding additional information in the video signal during the compression process H04N19/517, H04N19/68, H04N19/70 take precedence
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

A digital media encoder/decoder uses a flexible quantization technique that provides the ability to vary quantization along various dimensions of the encoded digital media data, including spatial, frequency sub bands and color channels. The codec utilizes a signaling scheme to signal various permutations of flexible quantization combinations efficiently for primary usage scenarios. When a choice of quantizer is available, the codec efficiently encodes the current quantizer by defining a subset of quantizers and indexes the current quantizer from the set.

Claims

1. A digital media processing method comprising:

storing at least part of a bitstream representing compressed digital media data in a memory; and

decoding the at least part of the bitstream by applying quantization parameters to transform coefficients of blocks of the digital media data and applying inverse operations of a transform to the blocks of the digital media data, wherein the quantization parameters are selectively variable across at least a spatial dimension, a frequency dimension, and a color channel dimension, and wherein the selective variability across each of the dimensions is signaled by signals in the bitstream.

2. The method of claim 1, wherein the signals include a signal indicating whether the same one or more quantization parameters apply to an entire frame.

3. The method of claim 1, wherein the signals include a signal indicating whether the same one or more quantization parameters apply for DC coefficients and AC coefficients.

4. The method of claim 1, wherein the signals include a signal indicating whether the same one or more quantization parameters apply for luminance values and chrominance values.

5. The device of claim 1, wherein the signals include a signal indicating whether the same one or more quantization parameters apply to an entire frame.

6. The device of claim 1, wherein the signals include a signal indicating whether the same one or more quantization parameters apply for DC coefficients and AC coefficients.

7. The device of claim 1, wherein the signals include a signal indicating whether the same one or more quantization parameters apply for luminance values and chrominance values.

8. A digital media processing method comprising:

storing at least part of a bitstream representing compressed digital media data in a memory; and

decoding the at least part of the bitstream by applying quantization parameters to transform coefficients of blocks of the digital media data and applying inverse operations of a transform to the blocks of the digital media data, wherein the quantization parameters applied to luminance coefficients and to chrominance coefficients are selectively variable between at least a first chrominance quantization mode in which the chrominance coefficients use the same one or more quantization parameters as the luminance coefficients and a second chrominance quantization mode in which the chrominance coefficients use different quantization parameters than the luminance coefficients,

wherein quantization parameters applied to DC coefficients and to AC coefficients are selectively variable between at least a first frequency quantization mode in which the AC coefficients use the same one or more quantization parameters as the DC coefficients and a second frequency quantization mode in which the AC coefficients use different quantization parameters than the DC coefficients.

9. The method of claim 8, wherein the first chrominance quantization mode and the second chrominance quantization mode are signaled by a signal in the bitstream.

10. The method of claim 8, wherein the quantization parameters are selectively variable between at least a first spatial dimension quantization mode in which the same one or more quantization parameters are applied across all subdivisions of a frame and a second spatial dimension quantization mode in which subdivisions of the frame use different quantization parameters.

11. A picture decoder device comprising:

a data storage buffer for storing at least part of a bitstream of encoded picture data; and

a processor programmed to decode the at least part of the bitstream by applying quantization parameters to transform coefficients of blocks of the picture data according to quantization rules selectively applied according to signals in the bitstream and applying inverse operations of a transform to the blocks of the picture data, the signals in the bitstream including a first signal for determining how a quantization rule is to be applied across a spatial dimension, a second signal for determining how a quantization rule is to be applied across a frequency dimension, and a third signal for determining how a quantization rule is to be applied across a color channel dimension.

12. The device of claim 11, wherein the first signal indicates whether or not the same one or more quantization parameters apply to an entire frame.

13. The device of claim 11, wherein the second signal indicates whether or not the same one or more quantization parameters apply for DC coefficients and AC coefficients.

14. The device of claim 11, wherein the third signal indicates whether or not the same one or more quantization parameters apply for luminance values and chrominance values.

15. A picture decoder device comprising:

a data storage buffer for storing at least part of a bitstream of encoded picture data; and

a processor programmed to decode the at least part of the bitstream by applying quantization parameters to transform coefficients of the blocks of the picture data and applying inverse operations of a transform to the blocks of the picture data, wherein the quantization parameters applied to luminance coefficients and to chrominance coefficients are selectively variable between at least a first chroma quantization mode in which the chrominance coefficients use the same one or more quantization parameters as the luminance coefficients and a second chroma quantization mode in which the chrominance coefficients use different quantization parameters than the luminance coefficients,

wherein quantization parameters applied to DC coefficients and to AC coefficients are selectively variable between at least a first frequency quantization mode in which the AC coefficients use the same one or more quantization parameters as the DC coefficients and a second frequency quantization mode in which the AC coefficients use different quantization parameters than the DC coefficients.

16. The device of claim 15, wherein the first chroma quantization mode and the second chroma quantization mode are signaled by a signal in the bitstream.

17. The device of claim 15, wherein the quantization parameters are selectively variable between at least a first spatial dimension quantization mode in which the same one or more quantization parameters are applied across all portions of a frame and a second spatial dimension quantization mode in which different portions of the frame use different quantization parameters.

18. A digital media processing method comprising:

encoding digital media data by applying operations of a transform to blocks of the digital media data and applying quantization parameters to transform coefficients of the blocks of the digital media data, wherein the quantization parameters are selectively variable across at least a spatial dimension, a frequency dimension, and a color channel dimension, and

generating a bitstream of encoded digital media data, the bitstream further including signals for indicating how the quantization parameters were applied across the spatial dimension, the frequency dimension, and the color channel dimension.

19. The method of claim 18, wherein the signals include a signal indicating whether the same one or more quantization parameters apply to an entire frame.

20. The method of claim 18, wherein the signals include a signal indicating whether the same one or more quantization parameters apply for DC coefficients and AC coefficients.

21. The method of claim 18, wherein the signals include a signal indicating whether the same one or more quantization parameters apply for luminance values and chrominance values.

22. A picture encoder device comprising:

a data storage buffer for storing picture data to be encoded; and

a processor programmed to;

encode digital media data by applying operations of a transform to blocks of the digital media data and applying quantization parameters to transform coefficients of the blocks of the digital media data, wherein the quantization parameters are selectively variable across at least a spatial dimension, a frequency dimension, and a color channel dimension, and

generate a bitstream of encoded digital media data, the bitstream further including signals for indicating how the quantization parameters were applied across the spatial dimension, the frequency dimension, and the color channel dimension.