Methods and apparatus for adaptive loop filtering in video encoders and decoders

Patent metadata

Field	Value
Publication number	US-9681132-B2
Application number	US-201113988470-A
Country	US
Kind code	B2
Filing date	Nov 22, 2011
Priority date	Nov 24, 2010
Publication date	Jun 13, 2017
Grant date	Jun 13, 2017

Abstract

Various implementations provide filtering for pictures. Several such implementations attempt to reduce artifacts caused by over-filtering in a decoding loop. Certain of these implementations provide multiple loop filters, and select a loop filter based on the result of a deblocking 5 filter. One particular implementation applies a first filter to a picture to produce a filtered picture. A second filter is selected based on a difference between a value of a particular pixel in the filtered picture and a value of a corresponding pixel in the picture. The second filter is applied to the particular pixel in the filtered picture. In another implementation, a signal or signal structure includes 10 an indication of the second filter, an encoding of at least a portion of a picture, and an encoding of at least a portion of a subsequent picture that has been encoded based on the output of the second filter above.

First claim

The invention claimed is:

1. A method for encoding video data, comprising:
applying a first filter to at least a portion of a picture to produce a filtered picture, wherein the picture is generated based on prediction data and residual data as an initial reconstructed picture at an encoder;
selecting a second filter based on a difference between a value of a particular pixel in the filtered picture and a value of a corresponding pixel in the initial reconstructed picture, wherein the corresponding pixel in the initial reconstructed picture is filtered by the first filter to produce the particular pixel in the filtered picture, wherein the selecting includes:
classifying the particular pixel in the filtered picture based on the difference between the value of the particular pixel in the filtered picture and the value of the corresponding pixel in the initial reconstructed picture, and
selecting the second filter based on a result of the classifying;

applying the second filter to the particular pixel in the filtered picture to produce a twice-filtered pixel in a second filtered picture;
encoding parameters for the second filter into a bitstream at the encoder; and
encoding at least a portion of a subsequent picture, based on a version of the second filtered picture having the twice-filtered pixel.

2. The method of claim 1 wherein applying the second filter comprises:
filtering the particular pixel in the filtered picture to generate a first value;
filtering the corresponding pixel in the picture to generate a second value; and
generating a third value, corresponding to the particular pixel, in the second filtered picture, based on a weighted sum of the first value and the second value.

3. The method of claim 1 , wherein the particular pixel in the filtered picture is classified into a first category if the difference exceeds a threshold, and wherein the particular pixel in the filtered picture is classified into a second category if the difference does not exceed the threshold.

4. The method of claim 1 wherein the first filter comprises a de-artifacting filter.

5. The method of claim 4 wherein the de-artifacting filter comprises a deblocking filter.

6. The method of claim 5 wherein the deblocking filter is configured to increase filtering beyond an optimal point of the deblocking filter, thereby increasing the difference and impacting the selecting operation.

7. The method of claim 1 wherein selecting the second filter is further based on an indication of a number of pixels in the filtered picture that exhibit differences in value from values of corresponding pixels in the picture.

8. An apparatus for encoding video data, comprising at least a memory and one or more processors, the one or more processors configured for performing at least the following:
applying a first filter to at least a portion of a picture to produce a filtered picture, wherein the picture is generated based on prediction data and residual data as an initial reconstructed picture at an encoder;
selecting a second filter based on a difference between a value of a particular pixel in the filtered picture and a value of a corresponding pixel in the initial reconstructed picture, wherein the corresponding pixel in the initial reconstructed picture is filtered by the first filter to produce the particular pixel in the filtered picture, wherein the selecting includes:
classifying the particular pixel in the filtered picture based on the difference between the value of the particular pixel in the filtered picture and the value of the corresponding pixel in the initial reconstructed picture, and
selecting the second filter based on a result of the classifying;

applying the second filter to the particular pixel in the filtered picture to produce a twice-filtered pixel in a second filtered picture;
encoding parameters for the second filter into a bitstream at the encoder; and
encoding at least a portion of a subsequent picture based on a version of the second filtered picture having the twice-filtered pixel.

9. A non-transitory processor readable medium having stored thereon a signal structure, the signal structure comprising:
a first picture portion comprising an encoding of at least a portion of a picture, wherein the picture is generated based on prediction data and residual data as an initial reconstructed picture at an encoder or as an initial decoded picture at a decoder;
a second picture portion comprising an encoding of at least a portion of a subsequent picture, the portion of the subsequent picture having been encoded based on a version of the picture having a twice-filtered pixel,
wherein the twice-filtered pixel is produced by (i) using a first filter to filter a particular pixel in the picture to produce a filtered pixel and (ii) using a second filter to filter the filtered pixel to produce the twice-filtered pixel, wherein the second filter is selected based on a result of a classification of the filtered pixel that is based on a difference between a value of the particular pixel from the picture and a value of the filtered pixel; and
a filter indication portion comprising an encoding of parameters for the second filter.

10. The apparatus of claim 8 wherein the one or more processors are configured to:
filter the particular pixel in the filtered picture to generate a first value;
filter the corresponding pixel in the picture to generate a second value; and
generate a third value, corresponding to the particular pixel, in the second filtered picture based on a weighted sum of the first value and the second value.

11. The apparatus of claim 8 wherein the one or more processors are configured to:
classify the particular pixel in the filtered picture into a first category if the difference exceeds a threshold; and
classify the particular pixel in the filtered picture into a second category if the difference does not exceed the threshold.

12. The apparatus of claim 8 wherein the first filter comprises a de-artifacting filter.

13. The apparatus of claim 12 wherein the de-artifacting filter comprises a deblocking filter.

14. The apparatus of claim 13 wherein the deblocking filter is configured to increase filtering beyond an optimal point of the deblocking filter, thereby increasing the difference and impacting the selecting operation.

15. The apparatus of claim 8 wherein the one or more processors are configured to select the second filter based on an indication of a number of pixels in the filtered picture that exhibit differences in value from values of corresponding pixels in the picture.

16. A method for decoding video data, comprising:
applying a first filter to at least a portion of a picture to produce a filtered picture, wherein the picture is generated based on prediction data and residual data as an initial decoded picture at a decoder;
selecting a second filter based on a difference between a value of a particular pixel in the filtered picture and a value of a corresponding pixel in the initial decoded picture, wherein the corresponding pixel in the initial decoded picture is filtered by the first filter to produce the particular pixel in the filtered picture, wherein the selecting includes:
classifying the particular pixel in the filtered picture based on the difference between the value of the particular pixel in the filtered picture and the value of the corresponding pixel in the initial decoded picture, and
selecting the second filter based on a result of the classifying;

applying the second filter to the particular pixel in the filtered picture to produce a twice-fi

Assignees

• Chen Qian
• Zheng Yunfei
• Yin Peng
• Lu Xiaoan
• Sole Joel
• Thomson Licensing Dtv

Inventors

• Chen Qian
• Zheng Yunfei
• Yin Peng
• Lu Xiaoan
• Sole Joel

Classifications

• H04N19/51

  Motion estimation or motion compensation · CPC title

• H04N19/82

  involving filtering within a prediction loop · CPC title

• H04N19/00909 Primary

  Electricity · mapped topic

• H04N19/439

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

• H04N19/86 Primary

  involving reduction of coding artifacts, e.g. of blockiness · CPC title