Drift reduction for quality scalable video coding

Patent metadata

Field	Value
Publication number	US-9578325-B2
Application number	US-201113005665-A
Country	US
Kind code	B2
Filing date	Jan 13, 2011
Priority date	Jan 13, 2010
Publication date	Feb 21, 2017
Grant date	Feb 21, 2017

Abstract

This invention is a method of scalable video encoding and decoding. The scalable video encoding codes both a base layer and an enhanced layer having greater resolution and/or refresh rate. Upon decoding some enhanced layer pictures may be dropped to reach a best resolution and refresh rate within a target data rate. Upon encoding a key picture in at least one group of pictures forming the video is a combined base layer/enhanced layer key picture. Such a combined base layer/enhanced layer key picture cannot be dropped on decoding. This technique reduces drift in the decoder.

First claim

What is claimed is:

1. A method of scalable video data encoding comprising the steps of:
encoding a video in groups of pictures in a base layer resolution and refresh rate;
encoding the video in groups of pictures in an enhanced layer resolution and refresh rate, at least one of the enhanced layer resolution or refresh rate greater than the corresponding base layer resolution or refresh rate;
determining a key picture for each group of pictures;
forming a combined key picture from a corresponding base layer key picture and a corresponding enhanced layer key picture for at least one of the groups of pictures; and
substituting the combined key picture for the corresponding base layer key picture and the corresponding enhanced layer key picture in an output video data stream.

2. The method of scalable video data encoding of claim 1 , wherein:
said step of forming a combined key picture for at least one of the groups of pictures combines the corresponding base layer key picture and the corresponding enhanced layer key picture for each group of pictures.

3. The method of scalable video data encoding of claim 1 , wherein:
said step of forming a combined key picture for at least one of the groups of pictures combines the corresponding base layer key picture and the corresponding enhanced layer key picture for every N group of pictures, where N is a positive integer.

4. A method of scalable video data decoding comprising the steps of:
receiving a video in groups of pictures in a base layer resolution and refresh rate and in an enhancement layer resolution and refresh rate, at least one of the enhanced layer resolution or refresh rate greater than the corresponding base layer resolution or refresh rate, wherein a key picture for at least one of the groups of pictures is a combined base layer/enhanced layer key picture;
decoding each picture in each group of pictures of the base layer;
decoding each combined base layer/enhanced layer key picture for each group of pictures having a combined base layer/enhanced layer key picture; and
decoding selective pictures in each group of pictures of the enhanced layer to provide a best resolution and refresh rate within a target data rate.

5. The method of scalable video data decoding of claim 4 , wherein:
if a group of pictures includes a combined base layer/enhanced layer key picture and a next group of pictures does not include a combined base layer/enhanced layer key picture predicting the key picture of the next group of pictures of the base layer and predicting the key picture of the next group of pictures of the enhanced layer from the combined base layer/enhanced layer key picture of the prior group of pictures.

6. The method of scalable video data decoding of claim 4 , wherein:
if a group of pictures includes a combined base layer/enhanced layer key picture and a next group of pictures includes a combined base layer/enhanced layer key picture predicting the combined key picture of the next group of pictures from the combined base layer/enhanced layer key picture of the prior group of pictures.

Assignees

• Mathew Manu
• Kandpal Sumit
• Texas Instruments Inc

Inventors

• Mathew Manu
• Kandpal Sumit

Classifications

• H04N19/33

  in the spatial domain · CPC title

• H04N19/31

  in the temporal domain · CPC title

• H04N19/00448 Primary

  Electricity · mapped topic

• H04N19/66 Primary

  involving data partitioning, i.e. separation of data into packets or partitions according to importance · CPC title

• H04N19/36 Primary

  Scalability techniques involving formatting the layers as a function of picture distortion after decoding, e.g. signal-to-noise [SNR] scalability · CPC title