Quantization adjustments for DC shift artifacts

• US 20070248163A1
• Filed: 04/07/2006
• Published: 10/25/2007
• Est. Priority Date: 04/07/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Lin, Chih-Lung, Hsu, Pohsiang, Zuo, Xiping
• Granted Patent: US 8,503,536 B2
• Time in Patent Office: Days
• US Class Current: 375/240.200
• CPC Class Codes: H04N 19/117 Filters, e.g. for pre-processing or post-processing sub-band filter banks H04N19/635
• CPC Class Codes: H04N 19/124 Quantisation
• CPC Class Codes: H04N 19/14 Coding unit complexity, e.g. amount of activity or edge presence estimation H04N19/146 takes precedence
• CPC Class Codes: H04N 19/154 Measured or subjectively estimated visual quality after decoding, e.g. measurement of distortion use of rate-distortion criteria H04N19/147
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/18 the unit being a set of transform coefficients
• CPC Class Codes: H04N 19/192 the adaptation method, adaptation tool or adaptation type being iterative or recursive
• CPC Class Codes: H04N 19/59 involving spatial sub-sampling or interpolation, e.g. alteration of picture size or resolution
• CPC Class Codes: H04N 19/61 in combination with predictive coding
• CPC Class Codes: H04N 19/86 involving reduction of coding artifacts, e.g. of blockiness

Abstract

A video encoder identifies one or more portions of a video picture that contain DC shift blocks and adjusts quantization (e.g., by selecting a smaller quantization step size) to reduce contouring artifacts when the picture is reconstructed. The encoder can identify the portion(s) of the picture that contain DC shift blocks by identifying one or more gradient slope regions in the picture and analyzing quantization effects on DC coefficients in the gradient slope region(s). The encoder can select a coarser quantization step size for a high-texture picture portion.

Claims

1. In a video encoder, a method comprising:

computing plural transform coefficients for each of plural blocks in a video picture;

identifying one or more portions of the video picture that contain DC shift blocks;

adjusting quantization for the video picture to reduce contouring artifacts when the video picture is reconstructed; and

outputting encoded data for the video picture.

2. The method of claim 1 wherein the identifying the one or more portions of the video picture that contain DC shift blocks comprises:

identifying one or more gradient slope regions in the video picture; and

analyzing quantization effects on DC coefficients for one or more of the plural blocks in the one or more gradient slope regions.

3. The method of claim 2 wherein the video encoder analyzes a texture map to identify the one or more gradient slope regions in the video picture.

4. The method of claim 1 wherein the adjusting quantization for the video picture comprises reducing a quantization step size for one or more macroblocks containing one or more of the DC shift blocks.

5. The method of claim 1 wherein the quantization comprises bi-level differential quantization.

6. The method of claim 1 wherein the quantization comprises multi-level differential quantization.

7. The method of claim 1 wherein the identifying the one or more portions of the video picture that contain DC shift blocks includes identifying the DC shift blocks using DC coefficients of the plural blocks or mean sample values of the plural blocks, and wherein the quantization uses a uniform quantizer or a non-uniform quantizer.

8. The method of claim 1 wherein the identifying the one or more portions of the video picture that contain DC shift blocks includes checking a list of DC coefficient values with associated DC shift artifacts at a given quantization step size.

9. The method of claim 1 wherein the identifying the one or more portions of the video picture that contain DC shift blocks includes:

identifying the DC shift blocks;

identifying one or more gradient slope regions; and

for one of the DC shift blocks in one of the one or more gradient slope regions, checking if any neighbor block both is smooth and has a lower or higher DC coefficient value.

10. The method of claim 1 wherein the adjusting quantization further comprises increasing quantization for a high-texture portion of the video picture.

11. One or more computer-readable media having stored thereon computer executable instructions to cause a computer to perform the method of claim 1.

12. In a video encoder having a quantizer, a method comprising:

computing a DC coefficient for each of plural blocks in a video picture;

of the plural blocks, detecting one or more blocks having a DC coefficient that changes value after quantization and inverse quantization at a first quantization step size;

selecting a second quantization step size for at least one macroblock comprising at least one of the one or more blocks having a DC coefficient that changes value after quantization and inverse quantization at the first quantization step size, wherein the second quantization step size is smaller than the first quantization step size;

using the second quantization step size in quantization for the at least one macroblock; and

outputting encoded data for the video picture.

13. The method of claim 12 wherein the first quantization step size is a picture quantization step size.

14. The method of claim 12 wherein the first quantization step size is a smooth region quantization step size that is smaller than a picture quantization step size.

15. The method of claim 12 wherein the second quantization step size is selected during bi-level differential quantization.

16. The method of claim 12 wherein the second quantization step size is smaller than a third quantization step size used for smooth regions in the video picture.

17. One or more computer-readable media having stored thereon computer executable instructions to cause a computer to perform the method of claim 12.

18. In a video encoder, a method comprising:

detecting a gradient slope region in a video picture, the gradient slope region comprising one or more DC shift blocks, each of the one or more DC shift blocks having a DC coefficient that changes value after quantization and inverse quantization at a first quantization step size;

selecting a second quantization step size for at least one picture portion comprising at least one of the one or more DC shift blocks having a DC coefficient that changes value after quantization and inverse quantization at the first quantization step size, wherein the second quantization step size is smaller than the first quantization step size;

using the second quantization step size in quantization for the at least one picture portion; and

outputting encoded data for the video picture.

19. The method of claim 18 further comprising:

selecting a third quantization step size for a high-texture picture portion in the video picture, wherein the third quantization step size is greater than the first quantization step size; and

using the third quantization step size in quantization for the high-texture picture portion.

20. One or more computer-readable media having stored thereon computer executable instructions to cause a computer to perform the method of claim 18.