View-independent decoding for omnidirectional video
For omnidirectional video such as 360-degree Virtual Reality (360VR) video, a video system that support independent decoding of different views of the omnidirectional video is provided. A decoder for such a system can extract a specified part of a bitstream to decode a desired perspective/face/view of an omnidirectional image without decoding the entire image while suffering minimal or no loss in coding efficiency.
1. A method comprising:
receiving a bitstream comprising an encoded sequence of omnidirectional images, each omnidirectional image having a plurality of views, each view having a set of view-specific data for decoding the view;
receiving a selection of a view from the plurality of views of one of the omnidirectional images in the sequence;
for the one of the omnidirectional images in the sequence, decoding the set of view-specific data of the selected view of the one of the omnidirectional images and also decoding a part of at least a first other view of the plurality of views of the one of the omnidirectional images that is adjacent to the selected view without decoding at least a second other view of the plurality of views of the one of the omnidirectional images that is not adjacent to the selected view; and
providing the selected view of the omnidirectional image based on the decoded set of view-specific data for display.
2. The method of claim 1 , wherein each view of the plurality of views is coded as a set of slices or a set of tiles.
3. The method of claim 1 , wherein the set of view-specific data of each view is independently entropy encoded, wherein encoding a set of view-specific data of a view comprises resetting an entropy coding state.
4. The method of claim 1 , wherein the set of view-specific data of the selected view has no dependency to view-specific data of another view.
5. The method of claim 1 , wherein the set of view-specific data of the selected view comprises a referential element associated with a pixel block that references data outside of the pixel block, the method further comprising modifying the referential element when the referential element references data outside of the view, wherein the modified referential element references data within the view.
6. The method of claim 5 , wherein the referential element is selected from a list of candidates comprising motion vectors associated with neighboring pixel blocks.
7. The method of claim 1 further comprising:
when the omnidirectional image is a base layer picture in a prediction hierarchy:
decoding view-specific data of all views of the omnidirectional image; and
allowing a selection of a different view.
8. An electronic apparatus comprising:
a video decoding circuit capable of:
receiving a bitstream comprising an encoded sequence of omnidirectional images, each omnidirectional image having a plurality of views, each view having a set of view-specific data for decoding the view;
receiving a selection of a view from the plurality of views of one of the omnidirectional images in the sequence;
for the one of the omnidirectional images in the sequence, decoding the set of view-specific data of the selected view of the one of the omnidirectional images and also decoding a part of at least a first other view of the plurality of views of the one of the omnidirectional images that is adjacent to the selected view without decoding at least a second other view of the plurality of views of the one of the omnidirectional images that is not adjacent to the selected view; and
displaying the selected view of the omnidirectional image based on the decoded set of view-specific data; and
a user interface circuit capable of providing the selection of the view.
9. The electronic apparatus of claim 8 , wherein the set of view-specific data of the selected view comprises a referential element associated with a pixel block that references data outside of the pixel block, the method further comprising modifying the referential element when the referential element references data outside of the view, wherein the modified referential element references data within the view.
10. The electronic apparatus of claim 8 , wherein the video decoder circuit is further capable of:
when the omnidirectional image is a base layer picture in a prediction hierarchy:
decoding view-specific data of all views of the omnidirectional image; and
allowing a selection of a different view.