Techniques for processing multiplane images
A device, an apparatus and associated methods are provided. In one embodiment, the method comprises obtaining a multi-plane image (MPI) representation of a three dimensional (3D) scene. The MPI representation includes a plurality of slices of content from the 3D scene, each slice corresponding to a different depth relative to a position of a first virtual camera. Each slice is decomposed into regular tiles; and the orientation of each tile is determined.
1 . A method comprising:
receiving a multi-plane image (MPI) representation of a three-dimensional (3D) image, the MPI representation comprising a plurality of slices, each slice of the plurality of slices corresponding to a different depth relative to a first position;
decomposing each slice of the plurality of slices into a plurality of tiles;
determining a position of each tile of the plurality of tiles within each slice of the plurality of slices;
determining orientation information for selected tiles of the plurality of tiles within each slice of the plurality of slices, wherein the orientation information is relative to a position and depth of a tile within a respective slice;
generating a stacked representation of the plurality of slices, wherein each slice of the stacked representation comprises a plurality of oriented tiles, wherein each oriented tile is oriented according to respective orientation information; and
generating a merged image using the stacked representation, the merged image representing a view of the 3D image relative to a second position.
2 . The method of claim 1 , further comprising:
determining tiles of the plurality of tiles to be discarded;
discarding the tiles of the plurality of tiles determined to be discarded; and
retaining a remainder of tiles of the plurality of tiles that were not discarded, wherein the selected tiles comprise the retained tiles.
3 . The method of claim 1 , wherein the orientation information comprises a depth value corresponding to at least one corner of a respective tile.
4 . The method of claim 1 , wherein the orientation information comprises a depth value corresponding to a center of a tile.
5 . The method of claim 1 , wherein the orientation information comprises a slope value of a depth of a tile.
6 . The method of claim 1 , wherein the orientation information comprises a value of an angle of a tile relative to a plane of a respective slice.
7 . The method of claim 1 , wherein the orientation information comprises a value of an angle of a tile relative to a surface that is normal to a plane of a respective slice.
8 . The method of claim 1 , wherein the orientation information is configured to adjust each slice of the stacked representation to be extendable outside a plane of a respective slice of the MPI representation.
9 . The method of claim 1 , wherein determining the orientation information for selected tiles of the plurality of tiles within each slice of the plurality of slices comprises decoding the orientation information from a bitstream.
10 . The method of claim 1 , wherein determining the position of each tile of the plurality of tiles within each slice of the plurality of slices comprises decoding tile position information from a bitstream.
11 . A device comprising:
a processor configured to:
receive a multi-plane image (MPI) representation of a three-dimensional (3D) image, the MPI representation comprising a plurality of slices, each slice of the plurality of slices corresponding to a different depth relative to a first position;
decompose each slice of the plurality of slices into a plurality of tiles;
determine a position of each tile of the plurality of tiles within each slice of the plurality of slices;
determine orientation information for each tile of the plurality of tiles within each slice of the plurality of slices, wherein the orientation information is relative to a position and depth of a tile within a respective slice;
generate a stacked representation of the plurality of slices, wherein each slice of the stacked representation comprises a plurality of oriented tiles, wherein each oriented tile is oriented according to respective orientation information; and
generate a merged image using the stacked representation, the merged image representing a view of the 3D image relative to a second position.
12 . The device of claim 11 , the processor further configured to:
determine tiles of the plurality of tiles to be discarded;
discard tiles of the plurality of tiles determined to be discarded; and
retain a remainder of tiles of the plurality of tiles that were not discarded, wherein the selected tiles comprise the retained tiles.
13 . The device of claim 11 , wherein the orientation information comprises a depth value corresponding to at least one corner of a respective tile.
14 . The device of claim 11 , wherein the orientation information comprises a depth value corresponding to a center of a tile.
15 . The device of claim 11 , wherein the orientation information comprises a slope value of a depth of a tile.
16 . The device of claim 11 , wherein the orientation information comprises a value of an angle of a tile relative to a plane of a respective slice.
17 . The device of claim 11 , wherein the orientation information comprises a value of an angle of a tile relative to a surface that is normal to a plane of a respective slice.
18 . The device of claim 11 , wherein the orientation information is configured to adjust each slice of the stacked representation to be extendable outside a plane of a respective slice of the MPI representation.
19 . The device of claim 11 , wherein to determine the orientation information for selected tiles of the plurality of tiles within each slice of the plurality of slices comprises to decode the orientation information from a bitstream.
20 . The device of claim 11 , wherein to determine the position of each tile of the plurality of tiles within each slice of the plurality of slices comprises to decode tile position information from a bitstream.