Multidimensional quantization techniques for video coding/decoding systems

Patent metadata

Field	Value
Publication number	US-11153594-B2
Application number	US-201615250608-A
Country	US
Kind code	B2
Filing date	Aug 29, 2016
Priority date	Aug 29, 2016
Publication date	Oct 19, 2021
Grant date	Oct 19, 2021

Abstract

Video compression and decompression techniques are disclosed that provide improved bandwidth control for video compression and decompression systems. In particular, video coding and decoding techniques quantize input video in multiple dimensions. According to these techniques, pixel residuals may be generated from a comparison of an array of input data to an array of prediction data. The pixel residuals may be quantized in a first dimension. After the quantization, the quantized pixel residuals may be transformed to an array of transform coefficients. The transform coefficients may be quantized in a second dimension and entropy coded. Decoding techniques invert these processes. In still other embodiments, multiple quantizers may be provided upstream of the transform stage, either in parallel or in cascade, which provide greater flexibility to video coders to quantize data in different dimensions in an effort to balance the competing interest in compression efficiency and quality of reconstructed video.

First claim

We claim:

1. A video coding method, comprising:
generating motion prediction data by motion-compensation with motion vectors from decoded reference image data of a video sequence;
generating motion residual blocks from a comparison of input pixel blocks of the video sequence to the motion prediction data;
quantizing, with a first quantization type, the motion residual blocks according to respective first residual quantization parameters;
quantizing, with a second quantization type, the first quantized residual blocks according to respective second residual quantization parameters;
quantizing, with a third quantization type, the second quantized residual blocks according to respective third residual quantization parameters;
transforming the third quantized residual blocks to respective transform coefficient blocks;
entropy coding the transform coefficient blocks; and
outputting the entropy coded coefficient blocks and the corresponding residual block quantization parameters.

2. The video decoding method of claim 1 , wherein the first quantization type is a spatial scaler, and the spatial scaler is applied to chroma components of the residuals and not applied to a luma component of the residual.

3. The video decoding method of claim 1 , wherein the first quantization type is a spatial scaler, and the spatial scalar is only applied to a subset of the blocks of residual data within a frame and is not applied to other blocks of residual data within the frame.

4. The video decoding method of claim 1 , wherein:
the first quantization type is a spatial scaler,
the spatial scalar is only applied chroma components of residual data and not applied to a luma component of residual data in a first subset of blocks within a frame, and
the spatial scalar is applied to both chroma and luma components of residual data in a second subset of blocks within a frame.

5. The video decoding method of claim 1 , wherein the first quantization type is a precision quantizer, and the precision quantizer applies a linear quantizer uniformly to each pixel-domain residual value within a color component of a block.

6. The video decoding method of claim 1 , wherein the first quantization type is a precision quantizer, and the precision quantizer applies a non-linear quantizer to each pixel-domain residual value within a block.

7. The video decoding method of claim 1 , wherein the first quantization type is a precision quantizer, and quantization parameters corresponding to a block are predicted based on intensity values in the corresponding block of the prediction data.

8. A video coding method, comprising:
generating prediction data from decoded reference image data of a video sequence of frames,
generating blocks of pixel residuals from a comparison of an blocks of input data to corresponding blocks of the prediction data,
first quantizing the blocks of pixel residuals according to first quantization parameters and with a first type of quantization wherein the first quantization parameters indicate which blocks of a frame the first quantizing is applied to and which other blocks of the same frame the first quantizing is not applied to,
second quantizing the blocks of first quantized residuals according to second quantization parameters and with a second type of quantization different from the first type of quantization wherein the second quantization parameters indicate which blocks of the frame the second quantizing is applied to and which other blocks of the same frame the second quantizing is not applied to,
transforming the second quantized residuals into transform coefficients,
entropy coding the transform coefficients, and
outputting the entropy coded transform coefficients, the first quantization parameters, and the second quantization parameters as coded video data.

9. The video coding method of claim 8 , further comprising:
third quantizing the blocks of second quantized residuals according to third quantization parameters and with a third type of quantization different from both the first type of quantization and second type of quantization, and
outputting the third quantization parameters,
wherein the transforming the second quantized residuals includes transforming the third quantization of the second quantized residuals.

10. The video coding method of claim 8 , wherein one of the first quantization type or the second quantization type is spatial scaler for reducing a resolution of residuals.

11. The video decoding method of claim 10 , wherein the spatial scaler is applied to chroma components of the residuals and not applied to a luma component of the residual.

12. The video decoding method of claim 10 , wherein the spatial scalar is only applied to a subset of the blocks of residual data within a frame and is not applied to other blocks of residual data within the frame.

13. The video decoding method of claim 10 , wherein:
the spatial scalar is only applied to chroma components of residual data and not applied to a luma component of residual data in a first subset of blocks within a frame, and
the spatial scalar is applied to both chroma and luma components of residual data in a second subset of blocks within a frame.

14. The video decoding method of claim 8 , wherein one of the first quantization type or the second quantization type is inverse precision quantizer of residual values.

15. The video decoding method of claim 14 , wherein the precision quantizer applies a linear quantizer uniformly to each pixel-domain residual value within a color component of a block.

16. The video decoding method of claim 14 , wherein the precision quantizer applies a non-linear quantizer to each pixel-domain residual value within a block.

17. The video decoding method of claim 8 , wherein quantization parameters corresponding to a block are predicted based on intensity values in the corresponding block of prediction data.

18. A video decoding method, comprising:
decoding reference image data of a coded video sequence of frames received from a channel;
predicting blocks of prediction data from decoded reference image data;
entropy decoding coded video data of the coded video sequence, yielding blocks of transform coefficients and corresponding quantization parameters;
transforming the transform coefficients in a transform-domain into blocks of reconstructed residual data in a pixel-domain;
first dequantizing the blocks of reconstructed residual data into blocks of first dequantized residual data by inverting a first quantization type based on the corresponding quantization parameters wherein the first quantization parameters indicate which blocks of a frame the first quantizing is applied to and which other blocks of the same frame the first quantizing is not applied to;
second dequantizing the blocks of first dequantized residual data into second dequantized residual data by inverting a second quantization type based on the corresponding quantization parameters wherein the second quantization parameters indicate which blocks of the frame the second quantizing is applied to and which other blocks of the same frame the second quantizing is not applied to; and
generating reconstructed pixel values from the blocks of second dequantized residual data and from the prediction data.

19. The video decoding method of claim 18 , wherein one of the first quantization type or the second quantization type is spatial scaler for reducing a resolution of residuals.

20. The video decoding method of claim 19 , wherein the spatial scal

Assignees

• Apple Inc

Inventors

• Tourapis Alexandros
• Su Yeping
• Singer David
• Wu Hsi-Jung

Classifications

• H04N19/48

  using compressed domain processing techniques other than decoding, e.g. modification of transform coefficients, variable length coding [VLC] data or run-length data (motion estimation in a transform domain H04N19/547; processing of decoded motion vectors H04N19/513) · CPC title

• H04N19/439

  using cascaded computational arrangements for performing a single operation, e.g. filtering · CPC title

• H04N19/436

  using parallelised computational arrangements · CPC title

• H04N19/30

  using hierarchical techniques, e.g. scalability (H04N19/63 takes precedence) · CPC title

• H04N19/186

  the unit being a colour or a chrominance component · CPC title