US 9,807,424 B2Granted

Adaptive selection of region size for identification of samples in a transition zone for overlapped block motion compensation

AssigneeQUALCOMM Incorporated

InventorsLiwei Guo, Peisong Chen, Marta Karczewicz

Life of this patent

Estimating the term…

Loading patent…

Abstract

During a video encoding or decoding process, a predicted prediction block is generated for a CU. The CU may have two or more prediction units (PUs). A computing device selects a neighbor region size. After the computing device selects the neighbor region size, samples in a transition zone of the prediction block are identified. Samples associated with a first PU are in the transition zone if neighbor regions that contain the samples also contain samples associated with a second PU. Samples associated with the second PU may be in the transition zone if neighbor regions that contain the samples also contain samples associated with the first PU. The neighbor regions have the selected neighbor region size. A smoothing operation is then performed on the samples in the transition zone.

Claims25

Dependent claims

A method of decoding video data, the method comprising:

partitioning, by a video decoder, a coding unit (CU) of a picture of the video data into a first prediction unit (PU) and a second PU, wherein the first PU does not overlap the second PU;
receiving, by the video decoder, a syntax element indicating an adaptively selected region size;
for each respective sample of the first PU:
determining, by the video decoder, whether any sample in a respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU, wherein the respective 2-dimensional region of samples for the respective sample of the first PU is centered on the respective sample of the first PU and has the region size indicated by the syntax element;
and
identifying, by the video decoder, whether the respective sample of the first PU is in a transition zone, wherein the respective sample of the first PU is identified as being in the transition zone in response to determining a sample in the respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU and the respective sample of the first PU is identified as not being in the transition zone in response to determining no sample in the respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU;
for each respective sample identified as being in the transition zone:
generating, by the video decoder, two predictions for the respective sample using a motion vector of the first PU and a motion vector of the second PU respectively;
and
including, by the video decoder, in a prediction block of the CU, in place of the respective sample, a sample resulting from blending the two predictions for the respective sample;
and
using, by the video decoder, reconstructed residual data of the CU and the prediction block of the CU to generate a reconstructed sample block for the CU.

9 more elements

A computing device that decodes video data, the computing device comprising:

a non-transitory storage medium configured to store a reconstructed sample block for a coding unit (CU) of a picture of the video data;
and
a processor configured to:
partition the CU into a first prediction unit (PU) and a second PU, wherein the first PU does not overlap the second PU;
receive a syntax element indicating an adaptively selected region size;
for each respective sample of the first PU:
determine whether any sample in a respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU, wherein the respective 2-dimensional region of samples for the respective sample of the first PU is centered on the respective sample of the first PU and has the region size indicated by the syntax element;
identify whether the respective sample of the first PU is in a transition zone, wherein the respective sample of the first PU is identified as being in the transition zone in response to determining a sample in the respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU and the respective sample of the first PU is identified as not being in the transition zone in response to determining no sample in the respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU;
for each respective sample identified as being in the transition zone:
generate two predictions for the respective sample using a motion vector of the first PU and a motion vector of the second PU respectively;
and
include, in a prediction block of the CU, in place of the respective sample, a sample resulting from blending the two predictions for the respective sample;
and
use reconstructed residual data of the CU and the prediction block of the CU to generate the reconstructed sample block for the CU.

11 more elements

A method of decoding video data, the method comprising:

partitioning, by a video decoder, a coding unit (CU) of a picture of the video data into a first prediction unit (PU) and a second PU, wherein the first PU does not overlap the second PU;
determining, by the video decoder, based on a characteristic of the CU, a region size;
for each respective sample of the first PU:
determining, by the video decoder, whether any sample in a respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU, wherein the respective 2-dimensional region of samples for the respective sample of the first PU is centered on the respective sample of the first PU and has the determined region size;
identifying, by the video decoder, whether the respective sample of the first PU is in a transition zone, wherein the respective sample of the first PU is identified as being in the transition zone in response to determining a sample in the respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU and the respective sample of the first PU is identified as not being in the transition zone in response to determining no sample in the respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU;
for each respective sample identified as being in the transition zone:
generating, by the video decoder, two predictions for the respective sample using a motion vector of the first PU and a motion vector of the second PU respectively;
and
including, by the video decoder, in a prediction block of the CU, in place of the respective sample, a sample resulting from blending the two predictions for the respective sample;
and
using, by the video decoder, reconstructed residual data of the CU and the prediction block of the CU to generate a reconstructed sample block for the CU.

8 more elements

A computing device that decodes video data, the computing device comprising:

a non-transitory storage medium configured to store a reconstructed sample block for a coding unit (CU) of the video data;
and
a processor configured to:
partition a coding unit (CU) of a picture of the video data into a first prediction unit (PU) and a second PU, wherein the first PU does not overlap the second PU;
determine, based on a characteristic of the CU, a region size;
for each respective sample of the first PU:
determine whether any sample in a respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU, wherein the respective 2-dimensional region of samples for the respective sample of the first PU is centered on the respective sample of the first PU and has the determined region size;
identify whether the respective sample of the first PU is in a transition zone, wherein the respective sample of the first PU is identified as being in the transition zone in response to determining a sample in the respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU and the respective sample of the first PU is identified as not being in the transition zone in response to determining no sample in the respective 2-dimensional region of samples for the respective sample of the first PU is in the second PU;
for each respective sample identified as being in the transition zone:
generate two predictions for the respective sample using a motion vector of the first PU and a motion vector of the second PU respectively;
and
include, in a prediction block of the CU, in place of the respective sample, a sample resulting from blending the two predictions for the respective sample;
and
use reconstructed residual data of the CU and the prediction block of the CU to generate a reconstructed sample block for the CU.

11 more elements

Dependencies are inferred from the claim text.

Facts

Application: 13/311,834Filed 6 Dec 2011 · us provisional application
Priority: 10 Jan 2011
Published: US 2012/0177120 A112 Jul 2012
Granted: 31 Oct 2017
Term ends: Not estimatedThe source records don’t support an estimate.

Classification8 codes

H04N 19/86
Using pre-processing or post-processing specially adapted for video compression: involving reduction of coding artifacts, e.g. of blockiness
H04N 19/117
Characterised by the element, parameter or selection affected or controlled by the adaptive coding: filters, e.g. for pre-processing or…
H04N 19/127
Characterised by the element, parameter or selection affected or controlled by the adaptive coding: prioritisation of hardware or computational…
H04N 19/159
Assigned coding mode, i.e. the coding mode being predefined or preselected to be further used for selection of another element or parameter:…

Family

Application61/431,408Us provisional application · filed 10 Jan 2011
Application61/450,532Us provisional application · filed 8 Mar 2011
Application61/450,538Us provisional application · filed 8 Mar 2011
Application61/537,450Us provisional application · filed 21 Sep 2011
US 2012/0177120 A1Pre-grant publication · published 12 Jul 2012
US 9,807,424 B2This patent

From the US filing history. Foreign family members aren’t listed.