METHOD AND APPARATUS FOR REDUCED RESOLUTION PARTITIONING

• US 20090196517A1
• Filed: 08/16/2007
• Published: 08/06/2009
• Est. Priority Date: 08/25/2006
• Status: Active Grant

• Current Assignee: Interdigital Vc Holdings, Inc. (InterDigital, Inc.)
• Original Assignee: Thomson Licensing (Vantiva SA)
• Inventors: Divorra Escoda, Oscar, Yin, Peng
• Granted Patent: US 8,363,936 B2
• Time in Patent Office: Days
• US Class Current: 382/240
• CPC Class Codes: H04N 19/103 Selection of coding mode or of prediction mode
• CPC Class Codes: H04N 19/119 Adaptive subdivision aspects, e.g. subdivision of a picture into rectangular or non-rectangular coding blocks
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/46 Embedding additional information in the video signal during the compression process H04N19/517, H04N19/68, H04N19/70 take precedence
• CPC Class Codes: H04N 19/61 in combination with predictive coding
• CPC Class Codes: H04N 19/70 characterised by syntax aspects related to video coding, e.g. related to compression standards
• CPC Class Codes: H04N 19/96 Tree coding, e.g. quad-tree coding

Abstract

There are provided methods and apparatus for reduced resolution partitioning. An apparatus includes an encoder for encoding video data using adaptive tree-based frame partitioning, wherein partitions are obtained from a combination of top-down tree partitioning and bottom-up tree joining.

Claims

1. An apparatus, comprising:

an encoder for encoding an image by performing an initial partitioning of the image into a mid-sized set of image blocks, selecting some of the image blocks for further partitioning, and selecting others of the image blocks for merging, wherein the further partitioning and the merging use adaptive tree-based frame partitioning.

2. The apparatus of claim 1, wherein partitions are obtained from a combination of top-down tree partitioning and bottom-up tree joining.

3. The apparatus of claim 2, wherein said encoder uses a block size of an initial partition to generate a tree-based frame partition, the block size being any of available block sizes capable of being used for tree-based frame partitioning relating to a particular video encoding standard or recommendation.

4. The apparatus of claim 2, wherein said encoder uses the bottom-up tree joining to generate super-block partitions for the video data.

5. The apparatus of claim 4, wherein said encoder generates any of 32×

32, 16×

32, 32×

16 partitions from 16×

16 macroblocks using the bottom up tree joining.

6. The apparatus of claim 4, wherein said encoder optimizes motion data in order to be globally efficient over an entire super-block.

7. The apparatus of claim 2, wherein said encoder selectively enables or disables at least one of the top-down tree partitioning and the bottom-up tree joining using a high level syntax.

8. The apparatus of claim 7, wherein said encoder codes the high level syntax in at least one of a slice header level, a Supplemental Enhancement Information (SEI) level, a picture parameter set level, a sequence parameter set level, a network abstraction layer unit header level, a picture level, a slice level, and a macroblock level.

9. The apparatus of claim 7, wherein said encoder at least one of implicitly and explicitly signals local super-block related information in at least one of a resultant bitstream corresponding to the video data and side information corresponding thereto.

10. The apparatus of claim 9, wherein said encoder at least one of implicitly and explicitly signals the local super-block related information using at least one additional syntax element at a macroblock level.

11. The apparatus of claim 9, wherein the local super-block related information comprises at least one of a super-block enable indication, a super-block disable indication, a super-block type, a super-block size, and a partition direction for a rectangular super-block.

12. The apparatus of claim 9, wherein the local super-block related information is embedded into only one block of a super-block or is distributed among more than one block belonging to a same group of blocks that are comprised in the super-block.

13. The apparatus of claim 9, wherein in at least one block in a super-block, at least one of block coding modes, the local super-block related information and motion information is directly derived for a current block to be encoded in the super-block from at least one of the local super-block related information, block mode information and motion information, explicitly or implicitly embedded in at least one neighboring block.

14. The apparatus of claim 13, wherein derivation rules for deriving the at least one of the block coding modes, the local super-block related information and the motion information depend on a block scanning order.

15. The apparatus of claim 9, wherein a block scanning order is equal or different from an existing block scan order imposed by a video coding recommendation or a video coding standard.

16. The apparatus of claim 2, wherein said encoder embeds at least one local super-block partition related syntax element into direct prediction modes, such that super-block frame partitioning is implemented based on adaptive direct prediction modes that allow different possible predictors to be used for the direct prediction modes.

17. The apparatus of claim 16, wherein the at least one local super-block partition related syntax element operates as an adaptive predictor selector for direct prediction, and wherein at least one of motion information and a block mode type is directly predicted or derived based on neighboring macroblock data and super-block syntax information.

18. The apparatus of claim 2, where residual coding is adapted to handle the super-block size using at least one of a super-block adapted transform size and super-block adapted signaling of at least one of coded quantized coefficients and non-coded quantized coefficients.

19. A method, comprising:

encoding an image by performing an initial partitioning of the image into a mid-sized set of image blocks, selecting some of the image blocks for further partitioning, and selecting others of the image blocks for merging, wherein the further partitioning and the merging use adaptive tree-based frame partitioning.

20. The method of claim 19, wherein partitions are obtained from a combination of top-down tree partitioning and bottom-up tree joining.

21. The method of claim 20, wherein said encoding step uses a block size of an initial partition to generate a tree-based frame partition, the block size being any of available block sizes capable of being used for tree-based frame partitioning relating to a particular video encoding standard or recommendation.

22. The method of claim 20, wherein said encoding step uses the bottom-up tree joining to generate super-block partitions for the video data.

23. The method of claim 22, wherein said encoding step generates any of 32×

32, 16×

32, 32×

16 partitions from 16×

16 macroblocks using the bottom up tree joining.

24. The method of claim 22, wherein said encoding step optimizes motion data in order to be globally efficient over an entire super-block.

25. The method of claim 20, wherein said encoding step selectively enables or disables at least one of the top-down tree partitioning and the bottom-up tree joining using a high level syntax.

26. The method of claim 25, wherein said encoding step codes the high level syntax in at least one of a slice header level, a Supplemental Enhancement Information (SEI) level, a picture parameter set level, a sequence parameter set level, a network abstraction layer unit header level, a picture level, a slice level, and a macroblock level.

27. The method of claim 25, wherein said encoding step at least one of implicitly and explicitly signals local super-block related information in at least one of a resultant bitstream corresponding to the video data and side information corresponding thereto.

28. The method of claim 27, wherein said encoding step explicitly signals the local super-block related information using at least one additional syntax element at a macroblock level.

29. The method of claim 27, wherein the local super-block related information comprises at least one of a super-block enable indication, a super-block disable indication, a super-block type, a super-block size, and a partition direction for a rectangular super-block.

30. The method of claim 27, wherein the local super-block related information is embedded into only one block of a super-block or is distributed among more than one block belonging to a same group of blocks that are comprised in the super-block.

31. The method of claim 27, wherein in at least one block in a super-block, at least one of block coding modes, the local super-block related information and motion information is directly derived for a current block to be encoded in the super-block from at least one of the local super-block related information, block mode information and motion information, explicitly or implicitly embedded in at least one neighboring block.

32. The method of claim 31, wherein derivation rules for deriving the at least one of the block coding modes, the local super-block related information and the motion information depend on a block scanning order.

33. The method of claim 27, wherein a block scanning order is equal or different from an existing block scan order imposed by a video coding recommendation or a video coding standard.

34. The method of claim 20, wherein said encoding step embeds at least one local super-block partition related syntax element into direct prediction modes, such that super-block frame partitioning is implemented based on adaptive direct prediction modes that allow different possible predictors to be used for the direct prediction modes.

35. The method of claim 34, wherein the at least one local super-block partition related syntax element operates as an adaptive predictor selector for direct prediction, and wherein at least one of motion information and a block mode type is directly predicted or derived based on neighboring macroblock data and super-block syntax information.

36. The method of claim 20, where residual coding is adapted to handle the super-block size using at least one of a super-block adapted transform size, and super-block adapted signaling of at least one of coded quantized coefficients and non-coded quantized coefficients.

37. An apparatus, comprising:

a decoder for decoding an image by performing an initial partitioning of the image into a mid-sized set of image blocks, selecting some of the image blocks for further partitioning, and selecting others of the image blocks for merging, wherein the further partitioning and the merging use adaptive tree-based frame partitioning.

38. The apparatus of claim 37, wherein partitions are obtained from a combination of top-down tree partitioning and bottom-up tree joining.

39. The apparatus of claim 38, wherein said decoder uses a block size of an initial partition to generate a tree-based frame partition, the block size being any of available block sizes capable of being used for tree-based frame partitioning relating to a particular video decoding standard or recommendation.

40. The apparatus of claim 38, wherein said decoder uses the bottom-up tree joining to generate super-block partitions for the video data.

41. The apparatus of claim 40, wherein said decoder generates any of 32×

32, 16×

32, 32×

16 partitions from 16×

16 macroblocks using the bottom up tree joining.

42. The apparatus of claim 38, wherein said decoder decodes at least one high level syntax element to selectively enable or disable at least one of the top-down tree partitioning and the bottom-up tree joining.

43. The apparatus of claim 42, wherein said decoder decodes the high level syntax in at least one of a slice header level, a Supplemental Enhancement Information (SEI) level, a picture parameter set level, a sequence parameter set level, a network abstraction layer unit header level, a picture level, a slice level, and a macroblock level.

44. The apparatus of claim 42, wherein said decoder at least one of implicitly and explicitly decodes local super-block related information from at least one of a resultant bitstream corresponding to the video data and side information corresponding thereto.

45. The apparatus of claim 44, wherein said decoder at least one of implicitly and explicitly decodes the local super-block related information by decoding at least one additional syntax element at a macroblock level.

46. The apparatus of claim 44, wherein the local super-block related information comprises at least one of a super-block enable indication, a super-block disable indication, a super-block type, a super-block size, and a partition direction for a rectangular super-block.

47. The apparatus of claim 44, wherein the at least one additional syntax element is decoded from only one block of a super-block or from a distribution from among more than one block belonging to a same group of blocks that are comprised in the super-block.

48. The apparatus of claim 44, wherein in at least one block in a super-block, at least one of block coding modes, the local super-block related information and motion information is directly derived for a current block to be encoded in the super-block from at least one of the local super-block related information, block mode information and motion information, explicitly or implicitly embedded in at least one neighboring block.

49. The apparatus of claim 48, wherein derivation rules for deriving the at least one of the block coding modes, the local super-block related information and the motion information depend on a block scanning order.

50. The apparatus of claim 44, wherein a block scanning order is equal or different from an existing block scan order imposed by a video coding recommendation or a video coding standard.

51. The apparatus of claim 38, wherein said decoder decodes at least one local super-block partition related syntax element from direct prediction modes, such that super-block frame partitioning is implemented based on adaptive direct prediction modes that allow different possible predictors to be used for the direct prediction modes.

52. The apparatus of claim 51, wherein the at least one local super-block partition related syntax element operates as an adaptive predictor selector for direct prediction, and wherein at least one of motion information and a block mode type is directly predicted or derived based on neighboring macroblock data and super-block syntax information.

53. The apparatus of claim 38, where residual decoding is adapted to handle the super-block size using at least one of a super-block adapted transform size, and super-block adapted signaling of at least one of coded quantized coefficients and non-coded quantized coefficients.

70. The method of claim 39, where residual decoding is adapted to handle the super-block size using at least one of a super-block adapted transform size, and super-block adapted signaling of at least one of coded quantized coefficients and non-coded quantized coefficients.

54. A method, comprising:

decoding an image by performing an initial partitioning of the image into a mid-sized set of image blocks, selecting some of the image blocks for further partitioning, and selecting others of the image blocks for merging, wherein the further partitioning and the merging use adaptive tree-based frame partitioning.

55. The method of claim 54, wherein partitions are obtained from a combination of top-down tree partitioning and bottom-up tree joining.

56. The method of claim 55, wherein said decoding step uses a block size of an initial partition to generate a tree-based frame partition, the block size being any of available block sizes capable of being used for tree-based frame partitioning relating to a particular video decoding standard or recommendation.

57. The method of claim 55, wherein said decoding step uses the bottom-up tree joining to generate super-block partitions for the video data.

58. The method of claim 57, wherein said decoding step generates any of 32×

32, 16×

32, 32×

16 partitions from 16×

16 macroblocks using the bottom up tree joining.

59. The method of claim 55, wherein said decoding step decodes at least one high level syntax element to selectively enable or disable at least one of the top-down tree partitioning and the bottom-up tree joining.

60. The method of claim 59, wherein said decoding step decodes the high level syntax from at least one of a slice header level, a Supplemental Enhancement Information (SEI) level, a picture parameter set level, a sequence parameter set level, a network abstraction layer unit header level, a picture level, a slice level, and a macroblock level.

61. The method of claim 55, wherein said decoding step at least one of implicitly and explicitly decodes local super-block related information from at least one of a resultant bitstream corresponding to the video data and side information corresponding thereto.

62. The method of claim 61, wherein said decoding step at least one of implicitly and explicitly decodes the local super-block related information by decoding at least one additional syntax element at a macroblock level.

63. The method of claim 61, wherein the local super-block related information comprises at least one of a super-block enable indication, a super-block disable indication, a super-block type, a super-block size, and a partition direction for a rectangular super-block.

64. The method of claim 61, wherein the at least one additional syntax element is decoded from only one block of a super-block or from a distribution from among more than one block belonging to a same group of blocks that are comprised in the super-block.

65. The method of claim 61, wherein the at least one additional syntax element selectively signals whether the local super-block related information is to be inherited by at least one neighboring block with respect to a previously coded block corresponding to the local super-block related information, and wherein at least one of block coding modes and motion information of the at least one neighboring block is directly derived from the at least one additional syntax element to allow direct decoding of the at least one of the block coding modes and the motion information of the at least one neighboring block.

66. The method of claim 65, wherein derivation rules for deriving the at least one of the block coding modes, the local super-block related information and the motion information depend on a block scanning order.

67. The method of claim 61, wherein a block scanning order is equal or different from an existing block scan order imposed by a video coding recommendation or a video coding standard.

68. The method of claim 55, wherein said decoding step decodes at least one local super-block partition related syntax element from direct prediction modes, such that super-block frame partitioning is implemented based on adaptive direct prediction modes that allow different possible-predictors to be used for the direct prediction modes.

69. The method of claim 68, wherein the at least one local super-block partition related syntax element operates as an adaptive predictor selector for direct prediction, and wherein at least one of motion information and a block mode type is directly predicted or derived based on neighboring macroblock data and super-block syntax information.

71. A video signal structure for video encoding, comprising:

an image encoded by performing an initial partitioning of the image into a mid-sized set of image blocks, selecting some of the image blocks for further partitioning, and selecting others of the image blocks for merging.

72. A storage media having video signal data encoded thereupon, comprising:

an image encoded by performing an initial partitioning of the image into a mid-sized set of image blocks, selecting some of the image blocks for further partitioning, and selecting others of the image blocks for merging.