Method for random access and gradual picture refresh in video coding
View Patent ↗A Method, And Associated Apparatus for providing random access to, and gradual refresh of, encoded video sequences. Gradual decoder refresh is enabled through the use of isolated regions, flexible macroblock order, and turn-off of loop filter at slice boundaries. Mechanisms are also provided for reliable detection of random access operations and for the reliable signaling of leading frames and ODR (open decoder refresh) pictures.
1. A method of encoding a video sequence, comprising:
providing a first indication in an encoded bitstream by a video encoder, the first indication indicating whether or not all coded pictures at and subsequent to an intra-coded picture in display order can be correctly decoded when a decoding process is started from the intra-coded picture; and
encoding a first picture into the encoded bitstream by the video encoder, using motion-compensated prediction with reference to one or more previously coded reference pictures, the first picture having a display order prior to the intra-coded picture and an encoding order succeeding the intra-coded picture;
providing a second indication in the encoded bitstream by the video encoder corresponding to the first picture, which indication indicates whether or not the first coded picture can be correctly decoded when the decoding process is started from the intra-coded picture;
performing motion compensated prediction by the video encoder for at least part of a second picture with reference to the first coded picture.
2. A method according to claim 1 , wherein the first indication and/or the second indication is provided in NAL unit-type syntax.
3. A method according to claim 1 , wherein the first indication and/or the second indication is provided in a picture header.
4. A method according to claim 1 , wherein the first indication and/or the second indication is provided in a slice header.
5. A method according to claim 1 , wherein an indication of a random access point is provided using a sub-sequence identifier.
6. A method of encoding a video sequence, comprising:
encoding, into an encoded bitstream by a video encoder, a first picture into the encoded bitstream, using motion-compensated prediction with reference to one or more previously coded reference pictures, the first picture having a display order prior to an intra-coded picture and an encoding order succeeding the intra-coded picture;
providing an indication in the encoded bitstream corresponding to the first picture, which indication indicates if at least one of said one or more previously coded reference pictures precedes the intra coded picture in decoding order; and
performing motion compensated prediction by the video encoder for a second picture with reference to said first picture.
7. A method according to claim 6 , wherein the first indication is provided in NAL unit-type syntax.
8. A method according to claim 6 , wherein the first indication is provided in a picture header.
9. A method according to claim 6 , wherein the first indication is provided in a slice header.
10. A method according to claim 6 , wherein an indication of a random access point using a sub-sequence identifier.
11. A method of decoding an encoded bitstream, comprising:
retrieving, by a video decoder, from the encoded bitstream, a first indication corresponding to an intra-coded picture, the first indication indicating that all coded pictures at and subsequent to the intra-coded picture in display order can be correctly decoded when a decoding process is started from the intra-coded picture; and
based on the decoded first indication, decoding the encoded bitstream by the video decoder, the decoding starting from the intra-coded picture,
decoding from the encoded bitstream a second indication corresponding to a first coded picture, the second indication indicating whether or not the first coded picture can be correctly decoded when decoding is started from the intra-coded picture, and
discarding the first coded picture without decoding.
12. A method according to claim 11 , wherein the indication is retrieved from NAL unit-type syntax.
13. A method according to claim 11 , wherein the indication is retrieved from a picture header.
14. A method according to claim 11 , wherein the indication is retrieved from a slice header.
15. A method, comprising:
retrieving, by a video decoder, from the encoded bitstream, a first indication corresponding to an intra-coded picture at a random access location, the first indication indicating that all decoded coded pictures at and subsequent to the intra-coded picture at the random access location in display order can be correctly decoded when a decoding process is started from the intra-coded picture at the random access location; and
retrieving, from the encoded video sequence, information corresponding to a first picture encoded using motion-compensated prediction with reference to one or more previously coded reference pictures, the first picture having a display order prior to the intra-coded picture at the random access location and a decoding order succeeding the intra-coded picture at the random access location; and
if the retrieved information corresponding to the first picture indicates that the at least one of said one or more previously coded reference pictures precedes the intra-coded picture at the random access location in decoding order, discarding the encoded first picture without decoding; and continuing the decoding process with the encoded pictures succeeding the first picture in the decoding order.
16. A method of decoding an encoded bitstream, comprising:
retrieving, from the encoded bitstream by a video decoder, a first indication corresponding to a first coded picture, the first indication indicating whether or not at least a part of at least the first coded picture was encoded with reference to a picture preceding a coded picture in encoding order, said first coded picture having a decoding order succeeding the intra-coded picture, said first coded picture used for motion-compensated prediction for at least a second coded picture; and
starting decoding from the intra-coded picture,
decoding the intra-coded picture,
decoding the first indication,
based on the decoded first indication, determining whether or not to decode the first coded picture.
17. A method according to claim 16 , further comprising, if the first indication indicates that at least a part of the at least one picture is encoded with reference to a picture preceding the intra-coded picture in encoding order:
discarding the at least one picture without decoding; and
continuing the decoding process with encoded pictures succeeding the first picture in the decoding order.
18. A method according to claim 16 , wherein the first indication is retrieved from NAL unit type syntax.
19. A method according to claim 16 , wherein the first indication is retrieved from a picture header.
20. A method according to claim 16 , wherein the first indication is retrieved from a slice header.
21. A method according to claim 16 , wherein a random access location is determined by examining sub-sequence identifiers for encoded pictures.
22. A method according to claim 11 , further comprising continuing the decoding process with the encoded pictures succeeding the first coded picture in the decoding order.
23. A method according to claim 22 , the first coded picture being used for motion compensated prediction for at least a part of a second coded picture.
24. An encoder configured for:
providing a first indication in an encoded bitstream by a video encoder, the first indication indicating whether or not all coded pictures at and subsequent to an intra-coded picture in display order can be correctly decoded when a decoding process is started from the intra-coded picture; and
encoding a first picture into the encoded bitstream by the video encoder, using motion-compensated prediction with reference to one or more previously coded reference pictures, the first picture having a display order prior to the intra-coded picture and an encoding order succeeding the intra-coded picture;
providing a second indication in the encoded bitstream by the video encoder corresponding to the first picture, which indication indicates whether or not the first coded picture can be correctly decoded when the decoding process is started from the intra-coded picture;
performing motion compensated prediction by the video encoder for at least part of a second picture with reference to the first coded picture.
25. A decoder configured for:
retrieving, by a video decoder, from the encoded bitstream, a first indication corresponding to an intra-coded picture, the first indication indicating that all coded pictures at and subsequent to the intra-coded picture in display order can be correctly decoded when a decoding process is started from the intra-coded picture; and
based on the decoded first indication, decoding the encoded bitstream by the video decoder, the decoding starting from the intra-coded picture,
decoding from the encoded bitstream a second indication corresponding to a first coded picture, the second indication indicating whether or not the first coded picture can be correctly decoded when decoding is started from the intra-coded picture, and
discarding the first coded picture without decoding.