Video encoding method and video decoding method

Patent metadata

Field	Value
Publication number	US-11575935-B2
Application number	US-202016902059-A
Country	US
Kind code	B2
Filing date	Jun 15, 2020
Priority date	Jun 14, 2019
Publication date	Feb 7, 2023
Grant date	Feb 7, 2023

Abstract

A video encoding method of encoding a multi-view image including one or more basic view images and a plurality of reference view images includes determining a pruning order of the plurality of reference view images, acquiring a plurality of residual reference view images, by pruning the plurality of reference view images based on the one or more basic view images according to the pruning order, encoding the one or more basic view images and the plurality of residual reference view images, and outputting a bitstream including encoding information of the one or more basic view images and the plurality of residual reference view images.

First claim

What is claimed is:

1. A video encoding method of encoding a multi-view image including a plurality of view images, the video encoding method comprising:
determining a pruning order of each of the plurality of view images, each of the plurality of view images being determined as a basic view image, which is a root node on a pruning hierarchy, or a reference view image, which is not the root node on the pruning hierarchy;
acquiring a residual reference view image by pruning the reference view image with at least one view image according to the pruning order;
generating one or more packed images based on the basic view image and the residual reference view image;
encoding the one or more packed images; and
outputting a bitstream including encoding information of the one or more packed images and pruning order information of the plurality of view images,
wherein the pruning order information comprises root node information indicating whether a view image is the root node, and
wherein when the root node information indicates that the view image is not the root node, the pruning order information further comprises number information specifying a number of parent nodes that have higher pruning priority on the pruning hierarchy than the view image and identity information of each of the parent nodes.

2. The video encoding method according to claim 1 , wherein the pruning order of each of the plurality of reference view images is determined based on a number of pixels included in an overlapping area between a reference view image and a basic view image.

3. The video encoding method according to claim 2 , wherein whether a pixel included in the reference view image is included in the overlapping area or not is determined by comparing the reference view image with a warped basic view image obtained by warping the basic view image to a view of the reference view image, or by comparing a warped reference view image obtained by warping the reference view image to a view of the basic view image with the basic view image.

4. The video encoding method according to claim 1 ,
wherein the pruning order of each of the plurality of reference view images is determined based on image indices of the plurality of reference view images, and
wherein each of the image indices is an acquisition order of the plurality of reference view images.

5. The video encoding method according to claim 1 ,
wherein the pruning order of each of the plurality of reference view images is determined based on camera parameters of the plurality of reference view images, and
wherein the camera parameters comprise at least one of a position or an angle of a camera.

6. The video encoding method according to claim 1 , wherein the pruning order of each of the plurality of reference view images is determined by:
acquiring, from the plurality of reference view images, a plurality of first residual reference view images, each of which an overlapping area with the basic view image is removed;
warping the plurality of first residual view images to a view of the basic view image; and
determining the pruning order of each of the plurality of reference view images according to the number of pixels of each of the plurality of warped first residual reference view images.

7. The video encoding method according to claim 1 , wherein acquiring the residual reference view image comprises:
generating a residual view image by removing an overlapping area between the reference view image and the basic view image from the reference view image; and
generating the residual reference view image by removing an overlapping area between the residual view image and an additional residual reference view image from the residual view image, the additional residual reference view image being generated by removing an overlapping area between an additional reference view image and the basic view image from the additional reference view image,
wherein the additional reference view image has a higher pruning priority on the pruning hierarchy than the reference view image.

8. The video encoding method according to claim 1 ,
wherein the one or more packed images is generated by packing a plurality of distributed valid areas included in the residual reference view image, and
wherein packing information for a distributed valid area is further encoded into the bitstream.

9. A video decoding method of decoding a multi-view image, the video decoding method comprising:
acquiring a bitstream including encoding information of one or more packed images and pruning order information;
decoding the one or more packed images based on the encoding information; and
reconstructing a viewport image based on the one or more packed images,
wherein the pruning order information comprises root node information indicating whether a view image is the root node, and
wherein when the root node information indicates that the view image is not the root node, the pruning order information further comprises number information specifying a number of parent nodes that have higher pruning priority on a pruning hierarchy than the view image and identity information of each of the parent nodes.

10. The video decoding method according to claim 9 ,
wherein the bitstream further comprises camera parameters comprising at least one of a position and/or an angle of a camera.

11. The video decoding method according to claim 9 ,
wherein the bitstream further comprises packing information of a valid area included in the one or more packed images, and
wherein the packing information indicates whether the valid area is rotated or resized in the one or more packed images.

12. A non-transitory computer-readable recording medium including a bitstream in which a plurality of view images is encoded, the bitstream comprising:
encoding information of one or more packed images; and
pruning order information of the plurality of view images,
wherein each of the plurality of view images is determined as a basic view image, which is a root node on a pruning hierarchy, or a reference view image, which is not the root node on the pruning hierarchy,
wherein the one or more packed images is generated based on a basic view image and a residual reference view image,
wherein the residual view image is obtained by pruning the reference view image with at least one view image according to the pruning order,
wherein the pruning order information comprises root node information indicating whether a view image is the root node, and
wherein when the root node information indicates that the view image is not the root node, the pruning order information further comprises number information specifying a number of parent nodes that have higher pruning priority on the pruning hierarchy than the view image and identity information of each of the parent nodes.

Assignees

• Electronics & Telecommunications Res Inst
• Iucf Hyu

Inventors

• Shin Hong Chang
• Lee Gwang Soon
• Eum Ho Min
• Jeong Jun Young
• Yun Kug Jin
• Yun Jun Young
• Park Jong Il

Classifications

• H04N19/172

  the region being a picture, frame or field · CPC title

• H04N19/597 Primary

  specially adapted for multi-view video sequence encoding · CPC title

• H04N19/105

  Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction · CPC title