Method and system for processing multiview videos for view synthesis using side information
First Claim
1. A method for processing a plurality of multiview videos of a scene, in which each video is acquired by a corresponding camera arranged at a particular pose, and in which a view of each camera overlaps with the view of at least one other camera, comprising the steps of:
- obtaining side information for synthesizing a particular view of the multiview video, wherein the side information includes depth values;
synthesizing, using the side information, a synthesized multiview video from an input video, wherein the input video includes at least one of the multiview videos, and the synthesized multiview video corresponds to single pose different than the input video;
maintaining a reference picture list for a picture associated with a particular time instant and a particular one of the plurality of multiview videos, wherein the reference picture list indexes temporal reference pictures and spatial reference pictures of the plurality of acquired multiview videos and the synthesized reference pictures of the synthesized multiview video, and wherein where the temporal reference pictures are associated with a time instant that is different than the particular time instant and the spatial reference pictures are associated with the same particular time instant, and wherein each reference picture is assigned a unique picture order count; and
predicting each current frame of the plurality of multiview videos according to reference pictures indexed by the associated reference picture list.
1 Assignment
0 Petitions
Accused Products
Abstract
A method processes a multiview videos of a scene, in which each video is acquired by a corresponding camera arranged at a particular pose, and in which a view of each camera overlaps with the view of at least one other camera. Side information for synthesizing a particular view of the multiview video is obtained in either an encoder or decoder. A synthesized multiview video is synthesized from the of multiview videos and the side information. A reference picture list is maintained for each current frame of each of the multiview videos, the reference picture indexes temporal reference pictures and spatial reference pictures of the acquired multiview videos and the synthesized reference pictures of the synthesized multiview video. Each current frame of the multiview videos is predicted according to reference pictures indexed by the associated reference picture list.
-
Citations
20 Claims
-
1. A method for processing a plurality of multiview videos of a scene, in which each video is acquired by a corresponding camera arranged at a particular pose, and in which a view of each camera overlaps with the view of at least one other camera, comprising the steps of:
-
obtaining side information for synthesizing a particular view of the multiview video, wherein the side information includes depth values; synthesizing, using the side information, a synthesized multiview video from an input video, wherein the input video includes at least one of the multiview videos, and the synthesized multiview video corresponds to single pose different than the input video; maintaining a reference picture list for a picture associated with a particular time instant and a particular one of the plurality of multiview videos, wherein the reference picture list indexes temporal reference pictures and spatial reference pictures of the plurality of acquired multiview videos and the synthesized reference pictures of the synthesized multiview video, and wherein where the temporal reference pictures are associated with a time instant that is different than the particular time instant and the spatial reference pictures are associated with the same particular time instant, and wherein each reference picture is assigned a unique picture order count; and predicting each current frame of the plurality of multiview videos according to reference pictures indexed by the associated reference picture list. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19)
-
-
20. A system for processing a plurality of multiview videos of a scene, comprising:
-
a plurality of cameras, each camera configured to acquire a multiview video of a scene, each camera arranged at a particular pose, and in which a view of each camera overlaps with the view of at least one other camera; means for obtaining side information for synthesizing a particular view of the multiview video, wherein the side information includes depth values; means for synthesizing a synthesized multiview video from an input video using the side information, wherein the input video includes at least one of the multiview videos, and the synthesized multiview video corresponds to single pose different than the input video; a memory for maintaining a reference picture list for a picture associated with a particular time instant and a particular one of the plurality of multiview videos, wherein the reference picture list indexes temporal reference pictures and spatial reference pictures of the plurality of acquired multiview videos and the synthesized reference pictures of the synthesized multiview video, and wherein where the temporal reference pictures are associated with a time instant that is different than the particular time instant and the spatial reference pictures are associated with the same particular time instant, and wherein each reference picture is assigned a unique picture order count; and means for predicting each current frame of the plurality of multiview videos according to reference pictures indexed by the associated reference picture list.
-
Specification