Bidirectionally predicted pictures or video object planes for efficient and flexible video coding

• US 6,404,813 B1
• Filed: 03/27/1997
• Issued: 06/11/2002
• Est. Priority Date: 03/27/1997
• Status: Expired due to Term

• Current Assignee: AT&T Intellectual Property II LP (AT&T, Inc.)
• Original Assignee: AT&T Corporation (AT&T, Inc.)
• Inventors: Haskell, Barin Geoffry, Schmidt, Robert Louis, Puri, Atul
• Primary Examiner(s): Kelley, Chris
• Assistant Examiner(s): PHILIPPE, GIMS S
• Application Number:
  US08/827,142
• Time in Patent Office: 1,902 Days
• Field of Search: 348/409, 348/403, 348/415, 348/416, 348/420, 348/398, 348/390, 348/447, 348/408, 348/699, 358/433, 375/240, 375/240.12, 375/240.16, 375/240.15, 275/240.16, 382/243, 382/236
• US Class Current: 375/240.12
• CPC Class Codes: H04N 19/105 Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction
• CPC Class Codes: H04N 19/20 using video object coding
• CPC Class Codes: H04N 19/56 Motion estimation with initialisation of the vector search, e.g. estimating a good candidate to initiate a search
• CPC Class Codes: H04N 19/577 Motion compensation with bidirectional frame interpolation, i.e. using B-pictures
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

A method is provided for decoding a bit stream representing an image that has been encoded. The method includes the steps of: performing an entropy decoding of the bit stream to form a plurality of transform coefficents and a plurality of motion vectors; performing an inverse transformation on the plurality of transform coefficients to form a plurality of error blocks; determining a plurality of predicted blocks based on bidirectional motion estimation that employs the motion vectors, wherein the bidirectional motion estimation includes a direct prediction mode and a second prediction mode; and, adding the plurality of error blocks to the plurality of predicted blocks to form the image. The second prediction mode may include forward, backward, and interpolated prediction modes.

Claims

1. A method for decoding a block in a B-picture, the method comprising:

receiving an indication of a prediction mode used to encode a block, the prediction mode including one from a group including forward prediction mode, backward prediction mode, interpolation mode and direct prediction mode;

if the indication indicates the direct prediction mode,

determining a first predicting block in a decoded past picture using a first scaling of a received motion vector associated with a block in a decoded future picture, the block in the future picture co-located with the block in the B-picture,

determining a second predicting block in the future picture using a second scaling of the received motion vector; and

thereafter, predicting data for the content of the block of the B-picture based on the first predicting block and the second predicting block.

2. The method of claim 1 wherein the forward prediction mode, the backward prediction mode and the interpolation mode are performed in accordance with the MPEG-1 video standard.

3. The method of claim 1 wherein the location of the first predicting block in the past picture is obtained based on the first scaling of the motion vector and the location of the block in B-picture, and the location of the second predicting block in the future picture is obtained based on the second scaling of the motion vector and the location of the block in B-picture.

4. The method of claim 3 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is corrected using a received correction motion vector.

5. The method of claim 4 wherein the second scaling of the motion vector is corrected using the received correction motion vector.

6. The method of claim 4 wherein the predicted data block is corrected using a received prediction error block.

7. The method of claim 6 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is based on the temporal-reference difference between the past picture and the future picture, and the temporal-reference difference between the past picture and the B-picture.

8. The method of claim 7 wherein the first scaling of the motion vector is performed according to the equation:

9. The method of claim 8 wherein the second scaling of the motion vector is performed according to the equation:

10. A method for predicting the content for a block of a first B-picture and a block of a second B-picture in an encoded bitstream, the method comprising:

decoding an entire past picture and an entire future picture;

decoding a motion vector associated with a block of the future picture, the block of the future picture co-located with the each of the block of the first B-picture and the block of the second B-picture;

thereafter, for each of the block of the first B-picture and the block of the second B-picture;

determining a first predicting block in the past picture and a second predicting block in the future picture using a first and second scaling, respectively, of the motion vector; and

predicting data for the content of the block of the B-picture based on the first predicting block and the second predicting block.

11. The method of claim 10 wherein, for each of the block of the first B-picture and the block of the second B-picture, the location of the first predicting block in the past picture is obtained based on the first scaling of the motion vector and the location of the block in B-picture, and the location of the second predicting block in the future picture is obtained based on the second scaling of the motion vector and the location of the block in B-picture.

12. The method of claim 11 wherein, for each of the block of the first B-picture and the block of the second B-picture, at least one of the first scaling of the motion vector and the second scaling of the motion vector is corrected using a received correction motion vector.

13. The method of claim 12 wherein, for each of the block of the first B-picture and the block of the second B-picture, the second scaling of the motion vector is corrected using the received correction motion vector.

14. The method of claim 12 wherein, for each of the block of the first B-picture and the block of the second B-picture, the predicted data block is corrected using a received prediction error block.

15. The method of claim 14 wherein, for each of the block of the first B-picture and the block of the second B-picture, at least one of the first scaling of the motion vector and the second scaling of the motion vector is based on the temporal-reference difference between the past picture and the future picture, and the temporal-reference difference between the past picture and that B-picture.

16. The method of claim 15 wherein, for each of the block of the first B-picture and the block of the second B-picture, the first scaling of the motion vector is performed according to the equation:

17. The method of claim 16 wherein, for each of the block of the first B-picture and the block of the second B-picture, the second scaling of the motion vector is performed according to the equation:

18. A method for decoding content for a block of a B-picture in an encoded bitstream, the method comprising:

receiving a motion vector associated with a block in a future picture, the block in the future picture co-located with the block of the B-picture, the motion vector identifying the location of a block in a past picture based on the location of the block in the future picture;

receiving the content of each of the past picture and the future picture;

thereafter, determining the location of a first predicting block in the past picture based on a first scaling of the motion vector and the location of the block of the B-picture;

determining the location of a second predicting block in the future picture based on a second scaling of the motion vector and the location of the block of the B-picture; and

generating a combination of the content of the first predicting block and the content of the second predicting block; and

transmitting the generated combination as the entire decoded block.

19. The method of claim 18 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is corrected using a received correction motion vector.

20. The method of claim 19 wherein the second scaling of the motion vector is corrected using the received correction motion vector.

21. The method of claim 19 wherein the generated combination is corrected using a received prediction error block before transmission.

22. A method for predicting the content for a block of a B-picture in an encoded bitstream, the method comprising:

determining a first predicting block in a decoded past picture using a first scaling of a received motion vector associated with a block in a decoded future picture, the block in the future picture co-located with the block in the B-picture

determining a second predicting block in the future picture using a second scaling of the received motion vector; and

thereafter, predicting data for the content of the block of the B-picture based on the first predicting block and the second predicting block.

23. The method of claim 22 wherein the location of the first predicting block in the past picture is obtained based on the first scaling of the motion vector and the location of the block in B-picture, and the location of the second predicting block in the future picture is obtained based on the second scaling of the motion vector and the location of the block in B-picture.

24. The method of claim 23 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is corrected using a received correction motion vector.

25. The method of claim 24 wherein the second scaling of the motion vector is corrected using the received correction motion vector.

26. The method of claim 24 wherein the predicted data block is corrected using a received prediction error block.

27. The method of claim 26 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is based on the temporal-reference difference between the past picture and the future picture, and the temporal-reference difference between the past picture and the B-picture.

28. The method of claim 27 wherein the first scaling of the motion vector is performed according to the equation:

29. The method of claim 28 wherein the second scaling of the motion vector is performed according to the equation:

30. A method for decoding a block in a B-VOP, the method comprising:

receiving an indication of a prediction mode used to encode a block, the prediction mode including one from a group including forward prediction mode, backward prediction mode, interpolation mode and direct prediction mode;

if the indication indicates the direct prediction mode,

determining a first predicting block in a decoded past VOP using a first scaling of a received motion vector associated with a block in a decoded future VOP, the future VOP co-located with the block in the B-VOP;

determining a second predicting block in the decoded future VOP using a second scaling of the received motion vector; and

thereafter, predicting data for the content of the block of the B-VOP based on the first predicting block and the second predicting block.

31. The method of claim 30 wherein the forward prediction mode, the backward prediction mode and the interpolation mode are performed in accordance with the MPEG-1 video standard.

32. The method of claim 30 wherein the location of the first predicting block in the past VOP is obtained based on the first scaling of the motion vector and the location of the block in B-VOP, and the location of the second predicting block in the future VOP is obtained based on the second scaling of the motion vector and the location of the block in B-VOP.

33. The method of claim 32 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is corrected using a received correction motion vector.

34. The method of claim 33 wherein the second scaling of the motion vector is corrected using the received correction motion vector.

35. The method of claim 33 wherein the predicted data block is corrected using a received prediction error block.

36. The method of claim 35 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is based on the temporal-reference difference between the past VOP and the future VOP, and the temporal-reference difference between the past VOP and the B-VOP.

37. The method of claim 36 wherein the block lies within a macroblock that lies entirely within the B-VOP.

38. The method of claim 36 wherein the block lies within a macroblock that lies on the border of the B-VOP.

39. The method of claim 38 wherein the block lies entirely within the B-VOP.

40. The method of claim 38 wherein the block lies on the border of the B-VOP.

41. The method of claim 36 wherein the first scaling of the motion vector is performed according to the equation:

42. The method of claim 41 wherein the second scaling of the motion vector is performed according to the equation:

43. A method for predicting the content for a block of a first B-VOP and a block of a second B-VOP in an encoded bitstream, the method comprising:

decoding a past VOP and a future VOP;

decoding a motion vector associated with a block of the future VOP, the block of the future VOP co-located with the each of the block of the first B-VOP and the block of the second B-VOP;

thereafter, for each of the block of the first B-VOP and the block of the second B-VOP;

determining a first predicting block in the past VOP and a second predicting block in the future VOP using a first and second scaling, respectively, of the motion vector; and

predicting data for the content of the block of the B-VOP based on the first predicting block and the second predicting block.

44. The method of claim 43 wherein, for each of the block of the first B-VOP and the block of the second B-VOP, the location of the first predicting block in the past VOP is obtained based on the first scaling of the motion vector and the location of the block in that B-VOP, and the location of the second predicting block in the future VOP is obtained based on the second scaling of the motion vector and the location of the block in that B-VOP.

45. The method of claim 44 wherein, for each of the block of the first B-VOP and the block of the second B-VOP, at least one of the first scaling of the motion vector and the second scaling of the motion vector is corrected using a received correction motion vector.

46. The method of claim 45 wherein, for each of the block of the first B-VOP and the block of the second B-VOP, the second scaling of the motion vector is corrected using a received correction motion vector.

47. The method of claim 45 wherein, for each of the block of the first B-VOP and the block of the second B-VOP, the predicted data block is corrected using a received prediction error block.

48. The method of claim 47 wherein, for each of the block of the first B-VOP and the block of the second B-VOP, at least one of the first scaling of the motion vector and the second scaling of the motion vector is based on the temporal-reference difference between the past VOP and the future VOP, and the temporal-reference difference between the past VOP and that B-VOP.

49. The method of claim 48 wherein each of the block of the first B-VOP and the block of the second B-VOP lies within a macroblock that lies entirely within the B-VOP.

50. The method of claim 48 wherein each of the block of the first B-VOP and the block of the second B-VOP lies within a macroblock that lies on the border of the B-VOP.

51. The method of claim 50 wherein each of the block of the first B-VOP and the block of the second B-VOP lies entirely within the B-VOP.

52. The method of claim 50 wherein each of the block of the first B-VOP and the block of the second B-VOP lies on the border of the B-VOP.

53. The method of claim 48 wherein, for each of the block of the first B-VOP and the block of the second B-VOP, the first scaling of the motion vector is performed according to the equation:

54. The method of claim 53 wherein, for each of the block of the first B-VOP and the block of the second B-VOP, the second scaling of the motion vector is performed according to the equation:

55. A method for decoding content for a block of a B-VOP in an encoded bitstream, the method comprising:

receiving a motion vector associated with a block in a future VOP, the block in the future VOP co-located with the block of the B-VOP, the motion vector identifying the location of a block in a past VOP based on the location of the block in the future VOP;

receiving the content of each of the past VOP and the future VOP;

thereafter, determining the location of a first predicting block in the past VOP based on a first scaling of the motion vector and the location of the block of the B-VOP;

determining the location of a second predicting block in the future VOP based on a second scaling of the motion vector and the location of the block of the B-VOP; and

generating a combination of the content of the first predicting block and the content of the second predicting block; and

transmitting the generated combination as the entire decoded block.

56. The method of claim 55 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is corrected using a received correction motion vector.

57. The method of claim 56 wherein the second scaling of the motion vector is corrected using the received correction motion vector.

58. The method of claim 57 wherein the generated combination is corrected using a received prediction error block before transmission.

59. The method of claim 58 wherein the block lies within a macroblock that lies entirely within the B-VOP.

60. The method of claim 58 wherein the block lies within a macroblock that lies on the border of the B-VOP.

61. The method of claim 60 wherein the block lies entirely within the B-VOP.

62. The method of claim 60 wherein the block lies on the border of the B-VOP.

63. A method for predicting the content for a block of a B-VOP in an encoded bitstream, the method comprising:

determining a first predicting block in a decoded past VOP using a first scaling of a received motion vector associated with a block in a decoded future VOP, the block in the future VOP co-located with the block in the B-VOP;

determining a second predicting block in a decoded future VOP using a second scaling of the received motion vector; and

thereafter, predicting data for the content of the block of the B-VOP based on the first predicting block and the second predicting block.

64. The method of claim 63 wherein the location of the first predicting block in the past VOP is obtained based on the first scaling of the motion vector and the location of the block in B-VOP, and the location of the second predicting block in the future VOP is obtained based on the second scaling of the motion vector and the location of the block in B-VOP.

65. The method of claim 64 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is corrected using a received correction motion vector.

66. The method of claim 65 wherein the second scaling of the motion vector is corrected using the received correction motion vector.

67. The method of claim 65 wherein the predicted data block is corrected using a received prediction error block.

68. The method of claim 67 wherein at least one of the first scaling of the motion vector and the second scaling of the motion vector is based on the temporal-reference difference between the past VOP and the future VOP, and the temporal-reference difference between the past VOP and the B-VOP.

69. The method of claim 68 wherein the block lies within a macroblock that lies entirely within the B-VOP.

70. The method of claim 68 wherein the block lies within a macroblock that lies on the border of the B-VOP.

71. The method of claim 70 wherein the block lies entirely within the B-VOP.

72. The method of claim 70 wherein the block lies on the border of the B-VOP.

73. The method of claim 68 wherein the first scaling of the motion vector is performed according to the equation:

74. The method of claim 73 wherein the second scaling of the motion vector is performed according to the equation: