Selective prediction signal filtering

• US 9,344,729 B1
• Filed: 07/11/2012
• Issued: 05/17/2016
• Est. Priority Date: 07/11/2012
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Grange, Adrian W, Bankoski, James, Wilkins, Paul G, Xu, Yaowu
• Primary Examiner(s): Tran, Thai
• Assistant Examiner(s): BRANIFF, CHRISTOPHER
• Application Number:
  US13/546,402
• Time in Patent Office: 1,406 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: H04N 19/103 Selection of coding mode or of prediction mode
• 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 19/147 according to rate distortion criteria rate-distortion as a criterion for motion estimation H04N19/567
• CPC Class Codes: H04N 19/80 Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation H04N19/635, H04N19/86 take precedence
• CPC Class Codes: H04N 19/82 involving filtering within a prediction loop

Abstract

Disclosed herein are methods and apparatuses for selective prediction signal filtering. One aspect of the disclosed implementations is a method for encoding a frame of a video stream including determining a first performance measurement for a first set of prediction samples identified for a group of pixels using a first prediction mode, generating a filtered set of prediction samples by applying a filter to a second set of prediction samples, wherein at least one of the filtered set of prediction samples or the second set of prediction samples are identified using a second prediction mode, determining a second performance measurement for the filtered set of prediction samples, generating, using a processor, a residual based on the filtered set of prediction samples and the group of pixels if the second performance measurement exceeds the first performance measurement, and encoding the frame using the residual.

Claims

1. A method for encoding a video signal having a frame comprising:

determining a first performance measurement for a first set of unfiltered prediction samples identified for a group of pixels of the frame using a first prediction mode;

generating a filtered set of prediction samples for the group of pixels by applying a filter to a second set of prediction samples, wherein at least one of the filtered set of prediction samples or the second set of prediction samples are identified using a second prediction mode;

determining a second performance measurement for the filtered set of prediction samples;

generating, using a processor, a residual based on the filtered set of prediction samples and the group of pixels if the second performance measurement exceeds the first performance measurement; and

encoding the frame using the residual wherein encoding the frame includes encoding a filter indicator indicating that the group of pixels was encoded using the filtered set of prediction samples and wherein the filter indicator includes a filter identifier that identifies the filter as at least one of a high-pass filter, a low-pass filter, a thresholded filter, a separable filter, or a non-separable filter and the filter indicator includes a filter state flag that is encoded for each block based on an encoding probability that is encoded at a frame level and is based on a number of blocks where the filter is enabled compared to a number of blocks where the filter is disabled.

2. The method of claim 1, wherein the performance measurements are rate distortion measurements.

3. The method of claim 1, wherein the first and second prediction modes are each an inter-prediction mode.

4. The method of claim 1, wherein the first set of unfiltered prediction samples is the same as the second set of prediction samples, and the first prediction mode is the same as the second prediction mode.

5. The method of claim 1, wherein the filter indicator includes a filter strength.

6. The method of claim 1, further comprising:

determining a third performance measurement for a set of pixels decoded from the encoded frame;

provisionally applying one or more filters to the set of pixels;

determining a fourth performance measurement for each filter provisionally applied to the selected set of pixels; and

generating a reconstruction of the frame using one or more selected filters, the selected filters selected based on comparisons indicating that the fourth performance measurement exceeds the third performance measurement.

7. The method of claim 6, wherein determining the fourth performance measurement for each filter includes determining a performance measurement for each of a plurality of strength settings for each of the provisionally applied filters such that each provisional application of a filter is associated with a plurality of performance measurements and generating the reconstruction includes selecting a strength setting for each of the selected filters based on the plurality of performance metrics.

8. The method of claim 1, wherein the filter is one of a sharpening filter or a blurring filter.

9. The method of claim 1, wherein the filter is a one dimensional filter with n taps applied to a target pixel in both a horizontal and vertical direction.

10. The method of claim 1, wherein applying the filter to the set of prediction samples includes applying the filter to a prediction sample in the set of prediction samples if a measurement of a filter dependent region of the prediction sample satisfies a threshold criteria.

11. The method of claim 10, wherein the measurement of the filter dependent region is a statistical variance and the threshold criteria is a pre-determined statistical variance.

12. The method of claim 1, wherein applying the filter to the second set of prediction samples includes applying the filter to pixels that are prospective prediction samples before the second set of prediction samples are identified using the second prediction mode.

13. The method of claim 12, wherein the second prediction mode is an intra-prediction mode.

14. The method of claim 1, wherein the first and second prediction modes identify prediction samples without using sub-pixel motion compensation.

15. The method of claim 1, wherein the filter is selected based on a motion vector identified by the second prediction mode.

16. The method of claim 15, wherein the filter blurs in a direction of the motion vector.

17. An apparatus for encoding a video signal having a frame comprising:

a memory; and

a processor configured to execute instructions stored in the at least one memory to;

apply a coding mode to reference frame data to produce unfiltered prediction data;

apply a filter to the unfiltered prediction data to produce filtered prediction data;

calculate a first value for a performance metric based on the unfiltered prediction data and current frame data;

calculate a second value for the performance metric based on the filtered prediction data and the current frame data; and

select the unfiltered prediction data or the filtered prediction data to encode a residual signal for the current frame data based on a comparison between the first value and the second value wherein encoding the residual signal includes encoding a filter indicator indicating that the residual signal was encoded using the filtered set of prediction samples and wherein the filter indicator includes a filter identifier that identifies the filter as at least one of a high-pass filter, a low-pass filter, a thresholded filter, a separable filter, or a non-separable filter and the filter indicator includes a filter state flag that is encoded for each block based on an encoding probability that is encoded at a frame level and is based on a number of blocks where the filter is enabled compared to a number of blocks where the filter is disabled.

18. A method for decoding an encoded video signal comprising:

decoding a prediction mode, a filter indicator, and a residual associated with a group of pixels of a frame of the encoded video signal, wherein the filter indicator includes a filter identifier that identifies the filter as at least one of a high-pass filter, a low-pass filter, a thresholded filter, a separable filter, or a non-separable filter and the filter indicator includes a filter state flag that is encoded for each block based on an encoding probability that is encoded at a frame level and is based on a number of blocks where the filter is enabled compared to a number of blocks where the filter is disabled;

identifying a set of prediction samples for decoding the group of pixels based on the prediction mode;

applying a filter to the identified set of prediction samples based on the filter indicator and the filter identifier to generate a filtered set of prediction samples; and

generating a reconstructed group of pixels using the filtered set of prediction samples and the residual.

19. The method of claim 18, further comprising:

decoding a loop filter indicator associated with the block from the encoded video signal; and

applying a loop filter to the reconstructed group of pixels based on the loop filter indicator to generate a group of pixels of an output video stream.

20. The method of claim 18, wherein the filter indicator is decoded using a prediction of the filter indicator based on at least one adjacent group of pixels to the group of pixels.