Apparatus and method for recovery of lost/damaged data in a bitstream of data based on compatibility of adjacent blocks of data

• US 6,263,108 B1
• Filed: 07/06/1998
• Issued: 07/17/2001
• Est. Priority Date: 10/23/1997
• Status: Expired due to Term

• Current Assignee: Sony Corporation (Sony Group Corp.), Sony Electronics Inc. (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.), Sony Electronics Inc. (Sony Group Corp.)
• Inventors: Carrig, James J., Fujimori, Yasuhiro, Ghosal, Sugata, Kondo, Tetsujiro
• Primary Examiner(s): Moore, David
• Assistant Examiner(s): Do, Anh Hong
• Application Number:
  US09/111,358
• Time in Patent Office: 1,107 Days
• Field of Search: 382/254, 382/266, 382/233, 382/232, 348/699, 371/43
• US Class Current: 382/232
• CPC Class Codes: H04N 19/59 involving spatial sub-sampling or interpolation, e.g. alteration of picture size or resolution
• CPC Class Codes: H04N 19/85 using pre-processing or post-processing specially adapted for video compression
• CPC Class Codes: H04N 19/88 involving rearrangement of data among different coding units, e.g. shuffling, interleaving, scrambling or permutation of pixel data or permutation of transform coefficient data among different blocks
• CPC Class Codes: H04N 19/89 involving methods or arrangements for detection of transmission errors at the decoder
• CPC Class Codes: H04N 19/895 in combination with error concealment
• CPC Class Codes: H04N 19/98 Adaptive-dynamic-range coding [ADRC]
• CPC Class Codes: H04N 21/426 Internal components of the client ; Characteristics thereof H04N5/44 takes precedence
• CPC Class Codes: H04N 5/21 Circuitry for suppressing or minimising disturbance, e.g. moiré or halo

Abstract

A system and method for receiving damage data in a bitstream of encoded data is disclosed. A plurality of candidate decodings for a plurality of blocks are generated. A measure is generated for each block based upon at least an image portion of the candidate decodings. Compatibility is determined between adjacent blocks of the plurality of blocks. A candidate decoding is selected based upon the measure and the compatibility of adjacent blocks. In one embodiment, this is used in the transmission of video signals over a potentially lossy communications channel.

Claims

1. A method for receiving lost/damaged data in a bitstream of encoded data comprising the steps of:

generating a plurality of candidate decodings for a group of related blocks within the bit stream;

generating at least one internal correlation measure for each block based upon at least an image portion of the candidate decodings to determine the intra-block correlation of adjacent pixels within the block;

determining an external measure compatibility between adjacent blocks of each block from the group of blocks to determine the inter-block compatibility;

combining the internal correlation measure of the block and the external compatibility measure of related blocks to generate a combined measure of the intra-block correlation and the inter-block compatibility; and

selecting a candidate decoding from the plurality of candidate decodings based upon the combined measure.

2. The method as set forth in claim 1, wherein the measure is a probability measure.

3. The method as set forth in claim 1, wherein the measure is a square error measure.

4. The method as set forth in claim 1, wherein the measure is a spatial correlation measure.

5. The method as set forth in claim 1, wherein the measure is a horizontal correlation measure.

6. The method as set forth in claim 1, wherein the measure is a vertical correlation measure.

7. The method as set forth in claim 1, wherein the measure is a diagonal correlation measure.

8. The method as set forth in claim 1, wherein the measure is a temporal activity measure.

9. A method for receiving lost/damaged data in a bitstream of encoded data comprising the steps of:

generating a plurality of candidate decodings for a plurality of blocks;

measuring for each block an internal smoothness measure of at least an image portion of the candidate decodings;

determining an external compatibility between adjacent blocks of the plurality of blocks;

combining the internal smoothness measure and the external compatibility measure of adjacent blocks to generate a combined measure; and

selecting a candidate decoding from the plurality of candidate decodings based upon the combined measure.

10. The method as set forth in claim 9, wherein measuring the smoothness comprises a Laplacian measurement.

11. The method as set forth in claim 9, wherein measuring smoothness further comprises:

normalizing x[i][j] that is defined as {tilde over (x)}[i][b] where x[j][j] is the decoded value at location [i,j]; and

computing smoothness using the x[i][j], Laplacian kernel L[m][n], and a block Laplacian value L_x, where L_x is determined according to the following;

$l<br><br>\left[i\right]<br><br>\left[j\right]=\underset{m=-1}{\overset{l}{\sum <br><br>}}<br><br>\underset{n=-1}{\overset{l}{\sum <br><br>}}<br><br>L<br><br>\left[m\right]<br><br>\left[n\right]<br><br>\tilde{x}<br><br>\left[i+m\right]<br><br>\left[j+n\right],\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K$$L_X=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br><br><br>l<br><br>\left[i\right]<br><br>\left[j\right]<br><br>}{K·<br><br>K},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K.$

12. The method as set forth in claim 11 where the {tilde over (x)}[i][j] is computed according to the following equation:

$\tilde{x}<br><br>\left[i\right]<br><br>\left[j\right]=\frac{x<br><br>\left[i\right]<br><br>\left[j\right]}{\sqrt{\underset{i,j}{\sum <br><br>}<br><br>{\left(x<br><br>\left[i\right]<br><br>\left[j\right]-X_{\mathrm{mean}}\right)}^2}},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K$$\mathrm{where},<br><br>X_{\mathrm{mean}}=\frac{\underset{i,j}{\sum <br><br>}<br><br>x<br><br>\left[i\right]<br><br>\left[j\right]}{K·<br><br>K},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K.$

13. The method as set forth in claim 11, wherein K is 8.

14. The method as set forth in claim 9, wherein measuring smoothness further comprises:

normalizing q[i][j], wherein q[i][j] is the Q code value at location [i,j] in an encoded image and q′

[i][j] is the normalized value of q[i][j];

computing smoothness using q′

[i][j], Laplacian kernel L[m][n], and a block Laplacian value L_q, where L_q equals $l<br><br>\left[i\right]<br><br>\left[j\right]=\underset{m=-1}{\overset{l}{\sum <br><br>}}<br><br>\underset{n=-1}{\overset{l}{\sum <br><br>}}<br><br>L<br><br>\left[m\right]<br><br>\left[n\right]<br><br>q^{\mathrm{\prime <br><br>}}<br><br>\left[i+m\right]<br><br>\left[j+n\right],\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right)$$L_q=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br><br><br>l<br><br>\left[i\right]<br><br>\left[j\right]<br><br>}{\left(K-2\right)·<br><br>\left(K-2\right)},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right).$

15. The method as set forth in claim 14, wherein q′

[i][j] is computed according to the following equation;

$q^{\mathrm{\prime <br><br>}}<br><br>\left[i\right]<br><br>\left[j\right]=\frac{q<br><br>\left[i\right]<br><br>\left[j\right]}{\sqrt{\underset{i,j}{\sum <br><br>}<br><br>{\left(q<br><br>\left[i\right]<br><br>\left[j\right]-Q_{\mathrm{mean}}\right)}^2}},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right)$$\mathrm{where},<br><br>Q_{\mathrm{mean}}=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>q<br><br>\left[i\right]<br><br>\left[j\right]}{\left(K-2\right)·<br><br>\left(K-2\right)},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right).$

16. The method as set forth in claim 14, wherein K is 8.

17. The method as set forth in claim 9, wherein the smoothness measure comprises a second order image surface property.

18. The method as set forth in claim 9, wherein the smoothness measure comprises an N order image surface property, where N>

1.

19. The method as set forth in claim 9, wherein measuring the smoothness further comprises:

determining if the image portion includes motion;

if the image portion includes motion,

computing a smoothness measurement for a first field of the image portion and a second field of the image portion, and

averaging the smoothness measurements together to generate the smoothness measure; and

if the image portion does not include motion, generating the smoothness measure for the image portion.

20. The method as set forth in claim 9, wherein determining further comprises evaluating spatial correlation to determine motion.

21. The method as set forth in claim 9, wherein determining a measure of compatibility between adjacent blocks further comprises determining the ratio between adjacent blocks and the total number of boundary pixels at the candidate block.

22. The method as set forth in claim 21, wherein a boundary pixel identified as compatible if it lies within a determined valid range of compatible boundary pixels.

23. The method as set forth in claim 9, further comprising generating a combined measure used to select a candidate decoding.

24. The method as set forth in claim 23, wherein the combined measure is determined according the following equation:

25. The method as set forth in claim 23, wherein the combined measure is determined according the following equation:

26. A processing system configured to receive lost/damaged data in a bitstream of encoded data comprising:

circuitry configured to generate a plurality of candidate decodings for a plurality of blocks;

circuitry configured to generate an internal correlation measure for each block based upon at least an image portion of the candidate decodings;

circuitry configured to determine an external compatibility measure between adjacent blocks of the plurality of blocks;

circuitry configured to combine the internal correlation measure and the external compatibility measure of adjacent blocks to generate a combined measure; and

circuitry configured to select a candidate decoding from the plurality of candidate decodings based upon the combined measure.

27. The processing system as set forth in claim 26, wherein the measure is a probability measure.

28. The processing system as set forth in claim 26, wherein the measure is a square error measure.

29. The processing system as set forth in claim 26, wherein the measure is a spatial correlation measure.

30. The processing system as set forth in claim 26, wherein the measure is a horizontal correlation measure.

31. The processing system as set forth in claim 26, wherein the measure is a vertical correlation measure.

32. The processing system as set forth in claim 26, wherein the measure is a diagonal correlation measure.

33. The processing system as set forth in claim 26, wherein the measure is a temporal activity measure.

34. A processing system for receiving lost/damaged data in a bitstream of encoded data comprising:

circuitry configured to generate a plurality of candidate decodings for a plurality of blocks;

circuitry configured to measure for each block the internal smoothness of at least an image portion of the candidate decodings;

circuitry configured to determine external compatibility measure between adjacent blocks of the plurality of blocks;

circuitry configured to combine the internal smoothness measure and the external compatibility measure of adjacent blocks to generate a combined measure; and

circuitry configured to select a candidate decoding from the plurality of candidate decodings based upon the combined measure.

35. The processing system as set forth in claim 34, wherein measuring the smoothness comprises a Laplacian measurement.

36. The processing system as set forth in claim 34, wherein measuring smoothness further comprises:

normalizing x[i][j] that is defined as {tilde over (x)}[i][j] where x[i][j] is the decoded value at location [i,j]; and

computing smoothness using the x[i][j], Laplacian kernel L[m][n], and a block Laplacian value L_x, where L_x is determined according to the following;

$l<br><br>\left[i\right]<br><br>\left[j\right]=\underset{m=-1}{\overset{l}{\sum <br><br>}}<br><br>\underset{n=-1}{\overset{l}{\sum <br><br>}}<br><br>L<br><br>\left[m\right]<br><br>\left[n\right]<br><br>\tilde{x}<br><br>\left[i+m\right]<br><br>\left[j+n\right],\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K$$L_X=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br><br><br>l<br><br>\left[i\right]<br><br>\left[j\right]<br><br>}{K·<br><br>K},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K.$

37. The processing system as set forth in claim 36, where the {tilde over (x)}[i][j] is computed according to the following equation:

$\tilde{x}<br><br>\left[i\right]<br><br>\left[j\right]=\frac{x<br><br>\left[i\right]<br><br>\left[j\right]}{\sqrt{\underset{i,j}{\sum <br><br>}<br><br>{\left(x<br><br>\left[i\right]<br><br>\left[j\right]-X_{\mathrm{mean}}\right)}^2}},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K$$\mathrm{where},<br><br>X_{\mathrm{mean}}=\frac{\underset{i,j}{\sum <br><br>}<br><br>x<br><br>\left[i\right]<br><br>\left[j\right]}{K·<br><br>K},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K.$

38. The processing system as set forth in claim 36, wherein K is 8.

39. The processing system as set forth in claim 34, wherein measuring smoothness further comprises:

normalizing q[i][j], wherein q[i][j] is the Q code value at location [i,j] in an encoded image and q′

[i][j] is the normalized value of q[i][j];

computing smoothness using q′

[i][j], Laplacian kernel L[m][n], and a block Laplacian value L_q, where L_q equals $l<br><br>\left[i\right]<br><br>\left[j\right]=\underset{m=-1}{\overset{l}{\sum <br><br>}}<br><br>\underset{n=-1}{\overset{l}{\sum <br><br>}}<br><br>L<br><br>\left[m\right]<br><br>\left[n\right]<br><br>q^{\mathrm{\prime <br><br>}}<br><br>\left[i+m\right]<br><br>\left[j+n\right],\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right)$$L_q=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br><br><br>l<br><br>\left[i\right]<br><br>\left[j\right]<br><br>}{\left(K-2\right)·<br><br>\left(K-2\right)},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right).$

40. The processing system as set forth in claim 39, wherein q′

[i][j] is computed according to the following equation;

$q^{\mathrm{\prime <br><br>}}<br><br>\left[i\right]<br><br>\left[j\right]=\frac{q<br><br>\left[i\right]<br><br>\left[j\right]}{\sqrt{\underset{i,j}{\sum <br><br>}<br><br>{\left(q<br><br>\left[i\right]<br><br>\left[j\right]-Q_{\mathrm{mean}}\right)}^2}},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right)$$\mathrm{where},<br><br>Q_{\mathrm{mean}}=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>q<br><br>\left[i\right]<br><br>\left[j\right]}{\left(K-2\right)·<br><br>\left(K-2\right)},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right).$

41. The processing system as set forth in claim 39, wherein K is 8.

42. The processing system as set forth in claim 34, wherein the smoothness measure comprises a second order image surface property.

43. The processing system as set forth in claim 34, wherein the smoothness measure comprises an N order image surface property, where N>

1.

44. The processing system as set forth in claim 34, wherein measuring the smoothness further comprises:

determining if the image portion includes motion;

if the image portion includes motion,

computing a smoothness measurement for a first field of the image portion and a second field of the image portion, and

averaging the smoothness measurements together to generate the smoothness measure; and

if the image portion does not include motion, generating the smoothness measure for the image portion.

45. The processing system as set forth in claim 34, wherein determining further comprises evaluating spatial correlation to determine motion.

46. The processing system as set forth in claim 34, wherein determining a measure of compatibility between adjacent blocks further comprises determining the ratio between adjacent blocks and the total number of boundary pixels at the candidate block.

47. The processing system as set forth in claim 46, wherein a boundary pixel identified as compatible if it lies within a determined valid range of compatible boundary pixels.

48. The processing system as set forth in claim 34, further comprising generating a combined measure used to select a candidate decoding.

49. The processing system as set forth in claim 48, wherein the combined measure is determined according the following equation:

50. The processing system as set forth in claim 48, wherein the combined measure is determined according the following equation:

51. A computer readable medium containing executable instructions which, when executed in a processing system, cause the system to receive lost/damaged data in a bitstream of encoded data, the system performing the steps comprising:

generating a plurality of candidate decodings for a plurality of blocks;

generating an internal correlation measure for each block based upon at least an image portion of the candidate decodings;

determining an external compatibility measure between adjacent blocks of the plurality of blocks;

combining the internal correlation measure and the external compatibility measure of adjacent blocks to generate a combined measure; and

selecting a candidate decoding from the plurality of candidate decodings based upon the combined measure.

52. The computer readable medium as set forth in claim 51, wherein the measure is a probability measure.

53. The computer readable medium as set forth in claim 51, wherein the measure is a square error measure.

54. The computer readable medium as set forth in claim 51, wherein the measure is a spatial correlation measure.

55. The computer readable medium as set forth in claim 51, wherein the measure is a horizontal correlation measure.

56. The computer readable medium as set forth in claim 51, wherein the measure is a vertical correlation measure.

57. The computer readable medium as set forth in claim 51, wherein the measure is a diagonal correlation measure.

58. The computer readable medium as set forth in claim 51, wherein the measure is a temporal activity measure.

59. A computer readable medium containing executable instructions which, when executed in a processing system, cause the system to receive lost/damaged data in a bitstream of encoded data, the system performing the steps comprising:

generating a plurality of candidate decodings for a plurality of blocks;

measuring for each block the internal smoothness of at least an image portion of the candidate decodings;

determining an external compatibility measure between adjacent blocks of the plurality of blocks;

combining the internal smoothness measure and the external compatibility measure of adjacent blocks to generate a combined measure; and

selecting a candidate decoding from the plurality of candidate decodings based upon the combined measure.

60. The computer readable medium as set forth in claim 59, wherein the step of measuring the smoothness comprises a Laplacian measurement.

61. The computer readable medium as set forth in claim 59, wherein the step of measuring smoothness comprises instructions, which when executed, perform the steps of:

normalizing x[i][j] that is defined as {tilde over (x)}[i][j] where x[i][j] is the decoded value at location [i,j]; and

computing smoothness using the x[i][j], Laplacian kernel L[m][n], and a block Laplacian value L_x, where L_x is determined according to the following;

$l<br><br>\left[i\right]<br><br>\left[j\right]=\underset{m=-1}{\overset{l}{\sum <br><br>}}<br><br>\underset{n=-1}{\overset{l}{\sum <br><br>}}<br><br>L<br><br>\left[m\right]<br><br>\left[n\right]<br><br>\tilde{x}<br><br>\left[i+m\right]<br><br>\left[j+n\right],\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K$$L_X=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br><br><br>l<br><br>\left[i\right]<br><br>\left[j\right]<br><br>}{K·<br><br>K},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K.$

62. The computer readable medium as set forth in claim 61, where the {tilde over (x)}[i][j] is computed according to the following equation:

$\tilde{x}<br><br>\left[i\right]<br><br>\left[j\right]=\frac{x<br><br>\left[i\right]<br><br>\left[j\right]}{\sqrt{\underset{i,j}{\sum <br><br>}<br><br>{\left(x<br><br>\left[i\right]<br><br>\left[j\right]-X_{\mathrm{mean}}\right)}^2}},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K$$\mathrm{where},\text{<br><br>}<br><br>X_{\mathrm{mean}}=\frac{\underset{i,j}{\sum <br><br>}<br><br>x<br><br>\left[i\right]<br><br>\left[j\right]}{K·<br><br>K},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K.$

63. The computer readable medium as set forth in claim 61, wherein K is 8.

64. The computer readable medium as set forth in claim 59, wherein the step of measuring smoothness comprises instructions which when executed perform the steps of:

normalizing q[i][j], wherein q[i][j] is the Q code value at location [i,j] in an encoded image and q′

[i][j] is the normalized value of q[i][j];

computing smoothness using q′

[i][j], Laplacian kernel L[m][n], and a block Laplacian value L_q, where L_q equals $l<br><br>\left[i\right]<br><br>\left[j\right]=\underset{m=-1}{\overset{l}{\sum <br><br>}}<br><br>\underset{n=-1}{\overset{l}{\sum <br><br>}}<br><br>L<br><br>\left[m\right]<br><br>\left[n\right]<br><br>q^{\mathrm{\prime <br><br>}}<br><br>\left[i+m\right]<br><br>\left[j+n\right],\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right)$$L_q=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br><br><br>l<br><br>\left[i\right]<br><br>\left[j\right]<br><br>}{\left(K-2\right)·<br><br>\left(K-2\right)},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right).$

65. The computer readable medium as set forth in claim 64, wherein q′

[i][j] is computed according to the following equation;

$q^{\mathrm{\prime <br><br>}}<br><br>\left[i\right]<br><br>\left[j\right]=\frac{q<br><br>\left[i\right]<br><br>\left[j\right]}{\sqrt{\underset{i,j}{\sum <br><br>}<br><br>{\left(q<br><br>\left[i\right]<br><br>\left[j\right]-Q_{\mathrm{mean}}\right)}^2}},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right)$$\mathrm{where},<br><br>Q_{\mathrm{mean}}=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>q<br><br>\left[i\right]<br><br>\left[j\right]}{\left(K-2\right)·<br><br>\left(K-2\right)},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right).$

66. The computer readable medium as set forth in claim 64, wherein K is 8.

67. The computer readable medium as set forth in claim 59, wherein the smoothness measure comprises a second order image surface property.

68. The computer readable medium as set forth in claim 59, wherein the smoothness measure comprises an N order image surface property, where N>

1.

69. The computer readable medium as set forth in claim 59, wherein the step of measuring the smoothness comprises instructions which when executed perform the steps of:

determining if the image portion includes motion;

if the image portion includes motion;

computing a smoothness measurement for a first field of the image portion and a second field of the image portion; and

averaging the smoothness measurements together to generate the smoothness measure; and

if the image portion does not include motion, generating the smoothness measure for the image portion.

70. The computer readable medium as set forth in claim 59, wherein the step of determining comprises instructions which when executed perform the step of evaluating spatial correlation to determine motion.

71. The computer readable medium as set forth in claim 59, wherein the step of determining a measure of compatibility between adjacent blocks comprises instructions which when executed perform the step of determining the ratio between adjacent blocks and the total number of boundary pixels at the candidate block.

72. The computer readable medium as set forth in claim 71, wherein a boundary pixel identified as compatible if it lies within a determined valid range of compatible boundary pixels.

73. The computer readable medium as set forth in claim 59, further comprising instructions which when executed perform the step of generating a combined measure used to select a candidate decoding.

74. The computer readable medium as set forth in claim 73, wherein the combined measure is determined according the following equation:

75. The computer readable medium as set forth in claim 73, wherein the combined measure is determined according the following equation:

76. An apparatus for receiving lost/damaged data in a bitstream of encoded data comprising:

means for generating a plurality of candidate decodings for a plurality of blocks;

means for measuring for each block the internal smoothness of at least an image portion of the candidate decodings;

means for determining an external compatibility measure between adjacent blocks of the plurality of blocks;

means for combining the internal smoothness measure and the external compatibility measure of adjacent blocks to generate a combined measure; and

means for selecting a candidate decoding from the plurality of candidate decodings based upon the combined measure.

77. The apparatus as set forth in claim 76, wherein the means for measuring the smoothness is configured to perform a Laplacian measurement.

78. The apparatus as set forth in claim 76, wherein the means for measuring smoothness comprises:

means for normalizing x[i][j] that is defined as {tilde over (x)}[i][j] where x[i][j] is the decoded value at location [i,j]; and

means for computing smoothness using the x[i][j], Laplacian kernel L[m][n], and a block Laplacian value L_x, where L_x is determined according to the following;

$l<br><br>\left[i\right]<br><br>\left[j\right]=\underset{m=-1}{\overset{l}{\sum <br><br>}}<br><br>\underset{n=-1}{\overset{l}{\sum <br><br>}}<br><br>L<br><br>\left[m\right]<br><br>\left[n\right]<br><br>\tilde{x}<br><br>\left[i+m\right]<br><br>\left[j+n\right],\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K$$L_X=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br><br><br>l<br><br>\left[i\right]<br><br>\left[j\right]<br><br>}{K·<br><br>K},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K.$

79. The apparatus as set forth in claim 78, where the {tilde over (x)}[i][j] is computed according to the following equation:

$\tilde{x}<br><br>\left[i\right]<br><br>\left[j\right]=\frac{x<br><br>\left[i\right]<br><br>\left[j\right]}{\sqrt{\underset{i,j}{\sum <br><br>}<br><br>{\left(x<br><br>\left[i\right]<br><br>\left[j\right]-X_{\mathrm{mean}}\right)}^2}},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K$$\mathrm{where},\text{<br><br>}<br><br>X_{\mathrm{mean}}=\frac{\underset{i,j}{\sum <br><br>}<br><br>x<br><br>\left[i\right]<br><br>\left[j\right]}{K·<br><br>K},\text{\hspace{1em}<br><br>}<br><br>0\le <br><br>i,j<<br><br>K.$

80. The apparatus as set forth in claim 76, wherein the means for measuring smoothness comprises:

means for normalizing q[i][j], wherein q[i][j] is the Q code value at location [i,j] in an encoded image and q′

[i][j] is the normalized value of q[i][j];

means for computing smoothness using q′

[i][j], Laplacian kernel L[m][n], and a block Laplacian value L_q, where L_q equals $l<br><br>\left[i\right]<br><br>\left[j\right]=\underset{m=-1}{\overset{l}{\sum <br><br>}}<br><br>\underset{n=-1}{\overset{l}{\sum <br><br>}}<br><br>L<br><br>\left[m\right]<br><br>\left[n\right]<br><br>q^{\mathrm{\prime <br><br>}}<br><br>\left[i+m\right]<br><br>\left[j+n\right],\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right)$$L_q=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br><br><br>l<br><br>\left[i\right]<br><br>\left[j\right]<br><br>}{\left(K-2\right)·<br><br>\left(K-2\right)},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right).$

81. The apparatus as set forth in claim 80, wherein q′

[i][j] is computed according to the following equation;

$q^{\mathrm{\prime <br><br>}}<br><br>\left[i\right]<br><br>\left[j\right]=\frac{q<br><br>\left[i\right]<br><br>\left[j\right]}{\sqrt{\underset{i,j}{\sum <br><br>}<br><br>{\left(q<br><br>\left[i\right]<br><br>\left[j\right]-Q_{\mathrm{mean}}\right)}^2}},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right)$$\mathrm{where},<br><br>Q_{\mathrm{mean}}=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>q<br><br>\left[i\right]<br><br>\left[j\right]}{\left(K-2\right)·<br><br>\left(K-2\right)},\text{\hspace{1em}<br><br>}<br><br>1\le <br><br>i,j<<br><br>\left(K-1\right).$

82. The apparatus as set forth in claim 80, wherein K is 8.