Multi-slice/tile encoder with overlapping spatial sections
View Patent ↗Roughly described, pictures are divided into multiple spatial sections to be encoded by multiple section encoders. To address discontinuities of compression decisions across section boundaries, the multiple section encoders encode overlapping regions in a picture.
1. A video encoding system for encoding an input video signal having a plurality of pictures including a particular picture, the particular picture having a plurality of unenlarged regions that includes first and second unenlarged regions that share at least one common boundary, comprising:
a plurality of encoders, each encoder implemented as program code in at least one processor that, when executed, causes the at least one processor to perform (a) encoding a different enlarged region in a plurality of enlarged regions of said particular picture and (b) outputting a different unenlarged region in said plurality of unenlarged regions for said particular picture, wherein each enlarged region that is encoded by one of said plurality of encoders has a greater area than the unenlarged region output by said one of said plurality of encoders,
wherein said plurality encoders include a first and second encoder, wherein encoding said different enlarged region in said plurality of enlarged regions comprises:
said first encoder receiving at least part of the particular picture and encoding a first enlarged region and outputting an encoded version of the first unenlarged region, wherein the first enlarged region has a larger area than said first unenlarged region;
said second encoder receiving at least part of the particular picture and encoding a second enlarged region and outputting an encoded version of the second unenlarged region; and
a multiplexer providing an output stream which includes, for the particular picture, the encoded version of the first unenlarged region from the first encoder and the encoded version of the second unenlarged region from the second encoder.
2. The system of claim 1 , wherein the second enlarged region encoded by the second hardware or software includes a region overlapping with the first enlarged region.
3. The system of claim 2 , wherein the first and second unenlarged regions each contain an integer number of encoders.
4. The system of claim 2 , wherein the particular picture further has a third enlarged region which partially overlaps with the second enlarged region,
wherein the second unenlarged region overlaps with a third unenlarged region, and
wherein the system further comprises a third encoder receiving at least part of the particular picture and encoding the third unenlarged region and outputting an encoded version of the third enlarged region as the third unenlarged region,
wherein the output stream provided by the multiplexer further includes the third unenlarged region from the third encoder.
5. The system of claim 1 , wherein the first and second unenlarged regions each define a respective slice of the particular picture.
6. The system of claim 1 , wherein the first and second unenlarged regions each define a respective integer number of tiles of the particular picture.
7. The system of claim 1 , wherein the first unenlarged region extends entirely across the particular picture horizontally.
8. The system of claim 1 , wherein the first unenlarged region extends horizontally across only a first horizontal range of the particular picture.
9. The system of claim 1 , wherein the first unenlarged region extends entirely across the particular picture vertically.
10. The system of claim 1 , wherein the first encoder encodes the first unenlarged region as a first spatial section of the particular picture, and encodes a first extension region as a second spatial section of the particular picture, wherein said first extension region corresponds to a remaining portion of said first enlarged region which is not in said first unenlarged region, and
wherein the encoding of the first unenlarged region takes into account encoding data selected from the group consisting of a motion vector from the first extension region, a coding mode decision from the first extension region, a choice from the first extension region of a reference picture, and a reconstructed pixel value from the first extension region.
11. The system of claim 1 , wherein the encoder encodes the first unenlarged region as a first spatial section of the particular picture, and encodes a first extension region as a second spatial section of the particular picture, wherein said first extension region corresponds to a remaining portion of said first enlarged region which is not in said first unenlarged region,
wherein the encoding of the first extension region takes into account encoding data selected from the group consisting of a motion vector from the first unenlarged region, a coding mode decision from the first unenlarged region, a choice from the first unenlarged region of a reference picture, and a reconstructed pixel value from the first unenlarged region.
12. The system of claim 1 , wherein the particular picture further has a third unenlarged region sharing at least a second common boundary with the first unenlarged region,
wherein the first enlarged region extends across the second common boundary into the third unenlarged region.
13. The system of claim 12 , wherein the first encoder encodes the first unenlarged region as a single spatial section of the particular picture.
14. The system of claim 12 , further comprising a third encoder receiving at least part of the particular picture and encoding the third unenlarged region and outputting an encoded version of the third unenlarged region,
wherein the output stream provided by the multiplexer further includes the encoded version of the third unenlarged region from the third encoder.
15. One or more non-transitory computer-readable storage mediums for encoding an input video signal having a plurality of pictures including a particular picture, the particular picture having a plurality of unenlarged regions that includes first and second unenlarged regions sharing at least one common boundary, which when executed by one or more processors, cause:
a plurality of hardware or software encoding units that each perform (a) encoding a different enlarged region in a plurality of enlarged regions of said particular picture and (b) outputting a different unenlarged region in said plurality of unenlarged regions for said particular picture, wherein each enlarged region that is encoded by one of said plurality of hardware or software encoding units has a greater area than the unenlarged region output by said one of said plurality of hardware or software encoding units,
wherein said plurality of hardware or software encoding units include a first and second hardware or software encoding units, wherein encoding said different enlarged region in said plurality of enlarged regions comprises:
said first hardware or software encoding unit encoding the first enlarged region, and
said second hardware or software encoding unit encoding the second enlarged region; and
providing an output stream which includes, for the particular picture, the encoded version of the first unenlarged region from the first hardware or software encoding unit and the encoded version of the second unenlarged region from the second hardware or software encoding unit.
16. The one or more non-transitory computer readable storage mediums of claim 15 , wherein the second enlarged region encoded by the second hardware or software includes a region overlapping with the first enlarged region.
17. The one or more non-transitory computer readable storage mediums of claim 16 , wherein the particular picture further has a third enlarged region which partially overlaps with the second enlarged region, wherein the second enlarged region encoded by the second hardware or software encoding unit overlaps with a third unenlarged region, and wherein execution of the one or more sequences of instructions further cause:
a third hardware or software encoding unit receiving at least part of the particular picture, encoding the third enlarged region, and outputting an encoded version of the third unenlarged region, wherein the third enlarged region comprises a greater area than the third unenlarged region, and
wherein the output stream further includes the encoded version of the third unenlarged region from the third encoding unit.
18. The one or more non-transitory computer readable storage mediums of claim 15 , wherein the first hardware or software encoding unit encodes the first unenlarged region as a single spatial section of the particular picture.
19. The one or more non-transitory computer readable storage mediums of claim 15 , wherein the first hardware or software encoding unit encodes the first enlarged region as a first spatial section of the particular picture, and encodes a first extension region as a second spatial section of the particular picture, wherein said first extension region corresponds to a remaining portion of said first enlarged region which is not in said first unenlarged region, and
wherein the encoding of the first unenlarged region takes into account encoding data selected from the group consisting of a motion vector from the first extension region, a coding mode decision from the first extension region, a choice from the first extension region of a reference picture, and a reconstructed pixel value from the first extension region.
20. The one or more non-transitory computer readable storage mediums of claim 15 , wherein the first hardware or software encoding unit encodes the first enlarged region as a first spatial section of the particular picture, and encodes a first extension region as a second spatial section of the particular picture, wherein said first extension region corresponds to a remaining portion of said first enlarged region which is not in said first unenlarged region, and
wherein the encoding of the first extension region takes into account encoding data selected from the group consisting of a motion vector from the first unenlarged region, a coding mode decision from the first unenlarged region, a choice from the first unenlarged region of a reference picture, and a reconstructed pixel value from the first unenlarged region.
21. The one or more non-transitory computer readable storage mediums of claim 15 , wherein the particular picture further has a third unenlarged region sharing at least a second common boundary with the first unenlarged region, and wherein the first enlarged region encoded by the first hardware or software encoding unit extends across the second common boundary into the third unenlarged region.
22. The one or more non-transitory computer readable storage mediums of claim 21 , wherein the first hardware or software encoding unit encodes the first unenlarged region as a single spatial section of the particular picture.