Apparatus and method for providing robust error recovery for errors that occur in a lossy transmission environment
First Claim
Patent Images
1. A method for recovering lost/damaged attribute data in a bitstream of encoded data comprising attribute data and encoded sample data, said method comprising the steps of:
- retrieving decoded neighboring data; and
estimating the lost/damaged attribute data using said encoded sample data, said decoded neighboring data, and available attribute data.
1 Assignment
0 Petitions
Accused Products
Abstract
A system and method for recovering lost/damaged attribute data in a bitstream of encoded data comprising attribute data and encoded sample data is disclosed. The decoded neighboring data is retrieved and lost/damaged attribute data is estimated using the encoded sample data, the decoded neighboring data, and available attribute data. In one embodiment, this is used in the transmission of video signals over a potentially lossy communications channel.
44 Citations
125 Claims
-
1. A method for recovering lost/damaged attribute data in a bitstream of encoded data comprising attribute data and encoded sample data, said method comprising the steps of:
-
retrieving decoded neighboring data; and estimating the lost/damaged attribute data using said encoded sample data, said decoded neighboring data, and available attribute data. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 95)
-
2. The method as set forth in claim 1, wherein the encoded data comprises a video signal.
-
3. The method as set forth in claim 1, wherein the encoded data comprises an audio signal.
-
4. The method as set forth in claim 1, wherein the lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, wherein said estimating the lost/damaged attribute data comprises the steps of:
-
gathering square error data of an estimation function with regard to said lost/damaged DR; and selecting a DR that minimizes said square error data of said estimation function.
-
-
5. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and wherein estimating comprises selecting an estimated value (DR'"'"') that minimizes the following function:
- ##EQU17## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and g() corresponds to a decoding operation.
-
6. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and wherein estimating comprises selecting an estimated value (DR'"'"') that minimizes the following function:
- ##EQU18## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
7. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and wherein estimating comprises selecting an estimated value (DR'"'"') that minimizes the following function:
- ##EQU19## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
8. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and said estimating comprises the following equation:
- ##EQU20## where Q is the quantization number of said samples, yi is a neighboring decoded sample value, and qi is an encoded value of said sample.
-
9. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and said estimating comprises the following equation:
- ##EQU21## where Q is the quantization number of said samples, yi is a neighboring decoded sample value, and qi is an encoded value of said sample.
-
10. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range (DR) of said samples, said method further estimating said the lost/damaged attribute data comprising the steps of:
-
gathering square error data of an estimation function with regard to said lost/damaged MIN; and selecting an estimated MIN value (MIN'"'"') that minimizes said square error data of said estimation function.
-
-
11. The method as set forth in claim 1, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and wherein estimating comprises selecting an estimated MIN value (MIN'"'"') that minimizes the following function:
- ##EQU22## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and g() corresponds to a decoding operation.
-
12. The method as set forth in claim 1, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and wherein estimating comprises selecting an estimated MIN value (MIN'"'"') that minimizes the following function:
- ##EQU23## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
13. The method as set forth in claim 1, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and wherein said estimating comprises selecting an estimated MIN value (MIN'"'"') that minimizes the following function:
- ##EQU24## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
14. The method as set forth in claim 1, further comprising clipping, wherein clipping comprises limiting estimated attribute data to the range of, where LQ and UQ respectively represent the lower and upper bounds of the range of possible attribute values for Q quantization bits.
-
15. The method as set forth in claim 14, wherein the lower and upper bounds of the clipping range of attribute values for Q quantization bits are relaxed outside the range of possible attribute values for suppressing visual degradation caused by instability of recovered attribute values.
-
16. The method as set forth in claim 15, wherein LQ and UQ are respectively adjusted to L'"'"'Q and U'"'"'Q according to the following equations:
-
space="preserve" listing-type="equation">L'"'"'.sub.Q =r.L.sub.Q-1 +(1-r).L.sub.Q
space="preserve" listing-type="equation">U'"'"'.sub.Q =(1-r).U.sub.Q +r.U.sub.Q+1wherein r is a relaxation constant, LQ-1 is the lower bound of the range of possible attribute values for Q-1 quantization bits, and UQ+1 is the upper bound of the range of possible attribute values for Q+1 quantization bits.
-
-
17. The method as set forth in claim 16, wherein r is equal to 0.5.
-
18. The method as set forth in claim 1, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and wherein said estimating comprises the following equation:
- ##EQU25## where Q is the quantization bit of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
19. The method as set forth in claim 1, wherein lost/damaged attribute data comprises said MIN as the samples, available attribute comprises said DR of said samples and wherein said comprises the following equation:
- ##EQU26## where Q is the quantization bit of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
20. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, the method further estimating said the lost/damaged attribute data comprising the steps of:
-
gathering square error data of an estimation function with regard to said lost/damaged DR and said lost/damaged MIN; and selecting a DR and a MIN that minimize said square error data of said estimation function.
-
-
21. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, and wherein estimating comprises selecting estimated values of DR'"'"' and MIN'"'"' that minimize the following function:
- ##EQU27## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and g() corresponds to a decoding operation.
-
22. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, and wherein estimating comprises selecting estimated values of DR'"'"' and MIN'"'"' that minimize the following function:
- ##EQU28## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
23. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, and wherein estimating comprises selecting estimated values of DR'"'"' and MIN'"'"' that minimize the following function:
- ##EQU29## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
24. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, and wherein estimating comprises selecting estimated values of DR'"'"' and MIN'"'"' that minimize the following functions:
- ##EQU30## where Q is the quantization bit number of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
25. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, and wherein estimating comprises selecting estimated values of DR'"'"' and MIN'"'"' that minimize the following functions:
- ##EQU31## where Q is the quantization bit number of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
26. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range value (DR) of said samples, and wherein said method further comprises estimating the lost/damaged attribute data by calculating a plurality of estimated data of the lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data;
- and selecting the median of said plurality of estimated data.
-
27. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples, available attribute comprises a dynamic range value (DR) of said samples and said estimating comprises selecting the median of a plurality of estimated data (MIN'"'"') according to the following equation:
- ##EQU32## where yi is a neighboring decoding sample value, qi is an encoded value of said sample, Q corresponds to a quantization number of said samples, and med() corresponds to median function.
-
28. The method as set forth in claim 24, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples, available attribute comprises a dynamic range value (DR) of said samples and said estimating comprises selecting the median of a plurality of estimated data (MIN'"'"') according to the following equation:
- ##EQU33## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, Q corresponds to a quantization number of said samples, and med() corresponds to a median function.
-
29. The method as set forth in claim 1, wherein estimating said lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data, further comprises the steps of:
-
selecting said decoded neighboring for said estimation according to said motion; and estimating lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data.
-
-
30. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and wherein estimating said the lost/damaged attribute data (DR'"'"') comprises the steps of:
-
estimating said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data; and clipping said estimated lost/damaged attribute data (DR'"'"') according to the following equation;
space="preserve" listing-type="equation">MIN+DR'"'"'≦
NUM.
-
-
31. The method as set forth in claim 30, wherein NUM is equal to 255.
-
32. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range value (DR) of said samples, and wherein said estimating said lost/damaged attribute data (MIN'"'"') comprises the steps of:
-
estimating said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data; and clipping said estimated lost/damaged attribute data (MIN'"'"') according to the following equation;
space="preserve" listing-type="equation">MIN'"'"'+DR≦
NUM.
-
-
33. The method as set forth in claim 32, wherein NUM is equal to 255.
-
34. The method as set forth in claim 1, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, and wherein said estimating said lost/damaged attribute data (DR'"'"' and MIN'"'"') further comprises the steps of:
-
estimating lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data; and clipping said estimated lost/damaged attribute data (DR'"'"' and MIN'"'"') according to the following equation;
space="preserve" listing-type="equation">MIN'"'"'+DR'"'"'≦
NUM.
-
-
35. The method as set forth in claim 34, wherein NUM is equal to 255.
-
36. The method as set forth in claim 1, wherein said estimating said lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data, further comprises the steps of:
- detecting correlation between said encoded data and corresponding neighboring decoded data;
selecting said neighboring decoded data for said estimation according to said correlation; and estimating said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data.
- detecting correlation between said encoded data and corresponding neighboring decoded data;
-
37. The method as set forth in claim 36, wherein lost/damaged attribute data comprising a dynamic range value (DR) of the samples, available attribute data comprising a minimum value (MIN) of said samples, qi is said encoded sample value, yi is said neighboring decoded sample value, and wherein detecting correlation comprises further steps of:
-
defining the range of {Lqi Uqi } corresponding to an encoded sample data qi, wherein Lqi is the lower bound corresponding to said qi and Uqi is the upper bound corresponding to said qi ; comparing a decoded neighboring data yi corresponding to said encoded data qi with said range {Lqi Uqi }; and deciding an introduction of said decoded neighboring data yi to said estimation step according to said comparison.
-
-
38. The method as set forth in claim 37, wherein lost/damaged attribute data comprising a dynamic range value (DR) of the samples, available attribute data comprising a minimum value (MIN) of said samples, and said comparison comprising further equations:
- ##EQU34## where yi is a neighboring decoded sample value, qi is an encoded sample value, Q is the number of quantization bits for said samples, m represents the maximum quantization code (2Q -1), LQ corresponds to the lower bound of Q bits, UQ corresponds to the upper bound of Q bits, and max and min respectively corresponds maximum and minimum functions.
-
39. The method as set forth in claim 1, wherein the encoded sample data and neighboring decoded data is grouped according to relative direction and wherein said estimating comprises:
-
generating directional estimates of the lost/damaged attribute data for each direction using corresponding encoded sample data and encoded neighboring data; weighting the directional estimates according relative amounts of correlation; and combining the weighted directional estimates to generate a combined estimate.
-
-
95. The computer readable medium as set forth in claim 94, 2 wherein r is equal to 0.5.
-
2. The method as set forth in claim 1, wherein the encoded data comprises a video signal.
-
-
40. A system comprising a processor configured to recover lost/damaged attribute data in a bitstream of encoded data comprising attribute data and encoded sample data, said processor configured to:
-
retrieve decoded neighboring data; and estimate the lost/damaged attribute data using said encoded sample data, said decoded neighboring data, and available attribute data. - View Dependent Claims (41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78)
-
41. The system as set forth in claim 40, wherein the encoded data comprises a video signal.
-
42. The system as set forth in claim 40, wherein the encoded data comprises an audio signal.
-
43. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and wherein said processor is configured to estimate the lost/damaged attribute data by gathering square error data of an estimation function with regard to said lost/damaged DR and selecting a DR that minimizes said square error data of said estimation function.
-
44. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and wherein said processor is configured to estimate by selecting an estimated value (DR'"'"') that minimizes the following function:
- ##EQU35## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and g() corresponds to a decoding operation.
-
45. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and said processor is configured to estimate by selecting an estimated value (DR'"'"') that minimizes the following function:
- ##EQU36## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
46. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and said processor is configured to estimate by selecting an estimated value (DR'"'"') that minimizes the following function:
- ##EQU37## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
47. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, said processor is configured to estimate using the following equation:
- ##EQU38## where Q is the quantization number of said samples, yi is a neighboring decoded sample value, and qi is an encoded value of said sample.
-
48. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, said processor is configured to estimate using the following equation:
- ##EQU39## where Q is the quantization number of said samples, yi is a neighboring decoded sample value, and qi is an encoded value of said sample.
-
49. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range (DR) of said samples, said system further configured to estimate the lost/damaged attribute data by gathering square error data of an estimation function with regard to said lost/damaged MIN, and selecting an estimated MIN value (MIN'"'"') that minimizes said square error data of said estimation function.
-
50. The system as set forth in claim 40, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and said processor is configured to estimate by selecting an estimated MIN value (MIN'"'"') that minimizes the following function:
- ##EQU40## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and g() corresponds to a decoding operation.
-
51. The system as set forth in claim 40, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and said processor is configured to select an estimated MIN value (MIN'"'"') that minimizes the following function:
- ##EQU41## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
52. The system as set forth in claim 40, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and said processor is configured to estimate by selecting an estimated MIN value (MIN'"'"') that minimizes the following function:
- ##EQU42## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a corresponding to a quantization number of said samples.
-
53. The system as set forth in claim 40, wherein the processor is further configured to clip by limiting estimated attribute data to the range of, where LQ and UQ respectively represent the lower and upper bounds of the range of possible attribute values for Q quantization bits.
-
54. The system as set forth in claim 53 , wherein the lower and upper bounds of the clipping range of attribute values for Q quantization bits are relaxed outside the range of possible attribute values for suppressing visual degradation caused by instability of recovered attribute values.
-
55. The system as set forth in claim 54, wherein LQ and UQ are respectively adjusted to L'"'"'Q and U'"'"'Q according to the following equations:
-
space="preserve" listing-type="equation">L'"'"'.sub.Q =r.L.sub.Q-1 +(1-r).L.sub.Q
space="preserve" listing-type="equation">U'"'"'.sub.Q =(1-r).U.sub.Q +r.U.sub.Q+1wherein r is a relaxation constant, LQ-1 is the lower bound of the range of possible attribute values for Q-1 quantization bits, and UQ+1 is the upper bound of the range of possible attribute values for Q+1 quantization bits.
-
-
56. The system as set forth in claim 55, wherein r is equal to 0.5.
-
57. The system as set forth in claim 40, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and the processor is configured to estimate using the following equation:
- ##EQU43## where Q is the quantization bit of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
58. The system as set forth in claim 40, wherein lost/damaged attribute data comprises said MIN as the samples, available attribute comprises said DR of said samples and the processor is configured to estimate using the following equation:
- ##EQU44## where Q is the quantization bit of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
59. The system as set forth in claim 40, wherein a lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, the processor further configured to estimate the lost/damaged attribute data by gathering square error data of an estimation function with regard to said lost/damaged DR and said lost/damaged MIN, and selecting a DR and a MIN that minimize said square error data of said estimation function.
-
60. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, and the processor is configured to estimate by selecting estimated values of DR'"'"' and MIN'"'"' that minimize the following function:
- ##EQU45## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and g() corresponds to a decoding operation.
-
61. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples and the processor is configured to select estimated values of DR'"'"' and MIN'"'"' that minimize the following function:
- ##EQU46## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
62. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples and the processor is configured to select estimated values of DR'"'"' and MIN'"'"' that minimize the following function:
- ##EQU47## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
63. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples and the processor is configured to select estimated values of DR'"'"' and MIN'"'"' that minimize the following functions:
- ##EQU48## where Q is the quantization bit number of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
64. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples and the processor is configured to select estimated values of DR'"'"' and MIN'"'"' that minimize the following functions:
- ##EQU49## where Q is the quantization bit number of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
65. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range value (DR) of said samples, and wherein said processor is further configured to estimate the lost/damaged attribute data by calculating a plurality of estimated data of the lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data, and selecting the median of said plurality of estimated data.
-
66. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples, available attribute comprises a dynamic range value (DR) of said samples and said processor is configured to estimate by selecting the median of a plurality of estimated data (MIN'"'"') according to the following equation:
- ##EQU50## where yi is a neighboring decoding sample value, qi is an encoded value of said sample, Q corresponds to a quantization number of said samples, and med() corresponds to median function.
-
67. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples, available attribute comprises a dynamic range value (DR) of said samples and the processor is configured to estimate by selecting the median of a plurality of estimated data (MIN'"'"') according to the following equation:
- ##EQU51## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, Q corresponds to a quantization number of said samples, and med() corresponds to a median function.
-
68. The system as set forth in claim 40, wherein the processor is configured to estimate said lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data, said processor configured to select said decoded neighboring for said estimation according to said motion, and estimate lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data.
-
69. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, the processor is configured to estimate the lost/damaged attribute data (DR'"'"') by estimating said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data, and clipping said estimated lost/damaged attribute data (DR'"'"') according to the following equation:
space="preserve" listing-type="equation">MIN+DR'"'"'≦
NUM.
-
70. The system as set forth in claim 69, wherein NUM is equal to 255.
-
71. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range value (DR) of said samples, the processor is configured to estimate the lost/damaged attribute data (MIN'"'"') by estimating said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data, and clipping said estimated lost/damaged attribute data (MIN'"'"') according to the following equation:
space="preserve" listing-type="equation">MIN'"'"'+DR≦
NUM.
-
72. The system as set forth in claim 71, wherein NUM is equal to 255.
-
73. The system as set forth in claim 40, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, and the processor is configured to estimate lost/damaged attribute data (DR'"'"' and MIN'"'"') by estimating lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data and clipping said estimated lost/damaged attribute data (DR'"'"' and MIN'"'"') according to the following equation:
space="preserve" listing-type="equation">MIN'"'"'+DR'"'"'≦
NUM.
-
74. The system as set forth in claim 73, wherein NUM is equal to 255.
-
75. The system as set forth in claim 40, wherein the processor is configured to estimate said lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data, said processor further configured to:
-
detect correlation between said encoded data and corresponding neighboring decoded data; select said neighboring decoded data for said estimation according to said correlation; and estimate said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data.
-
-
76. The system as set forth in claim 75, wherein lost/damaged attribute data comprising a dynamic range value (DR) of the samples, available attribute data comprising a minimum value (MIN) of said samples, qi is said encoded sample value, yi is said neighboring decoded sample value, and the processor is configured to detect correlation by
defining the range of {Lqi Uqi } corresponding to an encoded sample data qi, wherein Lqi is the lower bound corresponding to said qi and Uqi is the upper bound corresponding to said qi ; -
comparing a decoded neighboring data yi corresponding to said encoded data qi with said range {Lqi Uqi }; and deciding an introduction of said decoded neighboring data yi to said 12 estimation step according to said comparison.
-
-
77. The system as set forth in claim 76, wherein lost/damaged attribute data comprising a dynamic range value (DR) of the samples, available attribute data comprising a minimum value (MIN) of said samples, and said comparison comprising further equations:
- ##EQU52## where yi is a neighboring decoded sample value, qi is an encoded sample value, Q is the number of quantization bits for said samples, m represents the maximum quantization code (2Q -1), LQ corresponds to the lower bound of Q bits, UQ corresponds to the upper bound of Q bits, and max and min respectively corresponds maximum and minimum functions.
-
78. The system as set forth in claim 40, wherein the encoded sample data and neighboring decoded data is grouped according to relative direction and the processor is configured to estimate by generating directional estimates of the lost/damaged attribute data for each direction using corresponding encoded sample data and encoded neighboring data, weighting the directional estimates according relative amounts of correlation, and combining the weighted directional estimates to generate a combined estimate.
-
41. The system as set forth in claim 40, wherein the encoded data comprises a video signal.
-
-
79. A computer readable medium containing executable instructions which, when executed in a processing system, cause the system to recover lost/damaged attribute data in a bitstream of encoded data comprising attribute data and encoded sample data, the instructions causing the system to perform the following steps comprising:
-
retrieving decoded neighboring data; and estimating the lost/damaged attribute data using said encoded sample data, said decoded neighboring data, and available attribute data. - View Dependent Claims (80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117)
-
80. The computer readable medium as set forth in claim 79, wherein the encoded data comprises a video signal.
-
81. The computer readable medium as set forth in claim 79, wherein the encoded data comprises an audio signal.
-
82. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and wherein said estimating the lost/damaged attribute data comprises executable instructions, which when executed in the processing system cause the system to perform the following steps:
-
gathering square error data of an estimation function with regard to said lost/damaged DR; and selecting a DR that minimizes said square error data of said estimation function.
-
-
83. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and said estimating comprising executable instructions, which when executed in the processing system cause the system to select an estimated value (DR'"'"') that minimizes the following function:
- ##EQU53## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and g() corresponds to a decoding operation.
-
84. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and said estimating comprising executable instructions, which when executed in the processing system cause the system to select an estimated value (DR'"'"') that minimizes the following function:
- ##EQU54## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
85. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and said estimating comprising executable instructions, which when executed in the processing system cause the system to select an estimated value (DR'"'"') that minimizes the following function:
- ##EQU55## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
86. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and said estimating comprising executable instructions, which when executed in the processing system cause the system to perform said estimating using the following equation:
- ##EQU56## where Q is the quantization number of said samples, yi is a neighboring decoded sample value, and qi is an encoded value of said sample.
-
87. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, and said estimating comprising executable instructions, which when executed in the processing system cause the system to perform said estimating using the following equation:
- ##EQU57## where Q is the quantization number of said samples, yi is a neighboring decoded sample value, and qi is an encoded value of said sample.
-
88. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range (DR) of said samples, and said estimating comprising executable instructions, which when executed in the processing system cause the system to perform the following steps:
-
gathering square error data of an estimation function with regard to said lost/damaged MIN; and selecting an estimated MIN value (MIN'"'"') that minimizes said square error data of said estimation function.
-
-
89. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and said estimating comprising executable instructions, which when executed in the processing system cause the system to perform a step of selecting an estimated MIN value (MIN'"'"') that minimizes the following function:
- ##EQU58## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and g() corresponds to a decoding operation.
-
90. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and said estimating comprising executable instructions, which when executed in the processing system cause the system select an estimated MIN value (MIN'"'"') that minimizes the following function:
- ##EQU59## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
91. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and said estimating comprising executable instructions, which when executed in the processing system cause the system to perform the step selecting an estimated MIN value (MIN'"'"') that minimizes the following function:
- ##EQU60## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a corresponding to a quantization number of said samples.
-
92. The computer readable medium as set forth in claim 79, wherein clipping comprises executable instructions, which when executed in the processing system cause the system to limit estimated attribute data to the range of [LQ, UQ ], where LQ and UQ respectively represent the lower and upper bounds of the range of possible attribute values for Q quantization bits.
-
93. The computer readable medium as set forth in claim 92, wherein the lower and upper bounds of the clipping range of attribute values for Q quantization bits are relaxed outside the range of possible attribute values for suppressing visual degradation caused by instability of recovered attribute values.
-
94. The computer readable medium as set forth in claim 93, wherein LQ and UQ are respectively adjusted to L'"'"'Q and U'"'"'Q according to the following equations:
-
space="preserve" listing-type="equation">L'"'"'.sub.Q =r.L.sub.Q-1 +(1-r).L.sub.Q
space="preserve" listing-type="equation">U'"'"'.sub.Q= (1-r)U.sub.Q +r.U.sub.Q+1wherein r is a relaxation constant, LQ-1 is the lower bound of the range of possible attribute values for Q-1 quantization bits, and UQ+1 is the upper bound of the range of possible attribute values for Q+1 quantization bits.
-
-
96. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises said MIN of the samples, available attribute comprises said DR of said samples and said estimating comprising executable instructions, which when executed in the processing system cause the system to perform the step using the following equation:
- ##EQU61## where Q is the quantization bit of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
97. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises said MIN as the samples, available attribute comprises said DR of said samples and said estimating comprising executable instructions, which when executed in the processing system cause the system to perform the said estimating using the following equation:
- ##EQU62## where Q is the quantization bit of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
98. The computer readable medium as set forth in claim 79, wherein a lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, further comprising executable instructions, which when executed in the processing system cause the system the steps of:
-
gathering square error data of an estimation function with regard to said lost/damaged DR and said lost/damaged MIN; and selecting a DR and a MIN that minimize said square error data of said estimation function.
-
-
99. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples,
said estimating comprising executable instructions, which when executed in the processing system, cause the system to select estimated values of DR'"'"' and MIN'"'"' that minimize the following function: -
space="preserve" listing-type="equation">f(DR'"'"',MIN'"'"')=Σ
.sub.i (y.sub.i -g(q.sub.i,MIN'"'"',DR'"'"')).sup.2where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and g() corresponds to a decoding operation.
-
-
100. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples,
said estimating comprising executable instructions, which when executed in the processing system, cause the system to select estimated values of DR'"'"' and MIN'"'"' that minimize the following function: - ##EQU63## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
101. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples,
said estimating comprising executable instructions, which when executed in the processing system, cause the system to select estimated values of DR'"'"' and MIN'"'"' that minimize the following function: - ##EQU64## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, and Q corresponds to a quantization number of said samples.
-
102. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples,
said estimating comprising executable instructions, which when executed in the processing system, cause the system select estimated values of DR'"'"' and MIN'"'"' that minimize the following functions: - ##EQU65## where Q is the quantization bit number of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
103. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples,
said estimating comprising executable instructions, which when executed in the processing system, cause the system select estimated values of DR'"'"' and MIN'"'"' that minimize the following functions: - ##EQU66## where Q is the quantization bit number of said samples, yi is a neighboring decoded sample value, qi is an encoded value of said sample, and N corresponds to the number of terms used in the summation.
-
104. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range value (DR) of said samples, said medium comprising executable instructions, which when executed in the processing system, cause the system to estimate the lost/damaged attribute data comprising the steps of:
calculating a plurality of estimated data of the lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data; and
selecting the median of said plurality of estimated data.
-
105. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples, available attribute comprises a dynamic range value (DR) of said samples and said estimating comprising executable instructions, which when executed in the processing system, cause the system select the median of a plurality of estimated data (MIN'"'"') according to the following equation:
- ##EQU67## where yi is a neighboring decoding sample value, qi is an encoded value of said sample, Q corresponds to a quantization number of said samples, and med() corresponds to median function.
-
106. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples, available attribute comprises a dynamic range value (DR) of said samples and said estimating comprises selecting the median of a plurality of estimated data (MIN'"'"') according to the following equation:
- ##EQU68## where yi is a neighboring decoded sample value, qi is an encoded value of said sample, Q corresponds to a quantization number of said samples, and med() corresponds to a median function.
-
107. The computer readable medium as set forth in claim 79, wherein estimating said lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data comprises executable instructions, which when executed in the processing system, cause the system to perform the steps of:
-
selecting said decoded neighboring for said estimation according to said motion; and estimating lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data.
-
-
108. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, estimating said the lost/damaged attribute data (DR'"'"') comprising executable instructions, which when executed in the processing system, cause the system to perform the steps of:
-
estimating said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data; and clipping said estimated lost/damaged attribute data (DR'"'"') according to the following equation;
space="preserve" listing-type="equation">MIN+DR'"'"'≦
NUM.109.
-
-
109. The computer readable medium as set forth in claim 105, wherein NUM is equal to 255.
-
110. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range value (DR) of said samples, estimating said lost/damaged attribute data (MIN'"'"') comprising executable instructions, which when executed in the processing system, cause the system to perform the steps of:
-
estimating said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data; and clipping said estimated lost/damaged attribute data (MIN'"'"') according to the following equation;
space="preserve" listing-type="equation">MIN'"'"'+DR≦
NUM.
-
-
111. The computer readable medium as set forth in claim 110, wherein NUM is equal to 255.
-
112. The computer readable medium as set forth in claim 79, wherein lost/damaged attribute data comprises a dynamic range value (DR) and a minimum value (MIN) of the samples, estimating said lost/damaged attribute data (DR'"'"' and MIN'"'"') further comprising executable instructions, which when executed in the processing system, cause the system to perform the steps of:
-
estimating lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data; and clipping said estimated lost/damaged attribute data (DR'"'"' and MIN'"'"') according to the following equation;
space="preserve" listing-type="equation">MIN'"'"'+DR'"'"'≦
NUM.
-
-
113. The computer readable medium as set forth in claim 112, wherein NUM is equal to 255.
-
114. The computer readable medium as set forth in claim 79, wherein estimating said lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data, further comprising executable instructions, which when executed in the processing system, cause the system to perform the steps of:
-
detecting correlation between said encoded data and corresponding neighboring decoded data; selecting said neighboring decoded data for said estimation according to said correlation; and estimating said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data.
-
-
115. The computer readable medium as set forth in claim 114, wherein lost/damaged attribute data comprising a dynamic range value (DR) of the samples, available attribute data comprising a minimum value (MIN) of said samples, qi is said encoded sample value, yi is said neighboring decoded sample value, and said detecting correlation comprising executable instructions, which when executed in the processing system, cause the system to further perform the steps of:
-
defining the range of {Lqi Uqi } corresponding to an encoded sample data qi, wherein Lqi is the lower bound corresponding to said qi and Uqi is the upper bound corresponding to said qi ; comparing a decoded neighboring data yi corresponding to said encoded data qi with said range {Lqi Uqi }; and deciding an introduction of said decoded neighboring data yi to said estimation step according to said comparison.
-
-
116. The computer readable medium as set forth in claim 115, wherein lost/damaged attribute data comprising a dynamic range value (DR) of the samples, available attribute data comprising a minimum value (MIN) of said samples, and said instruction comprising the comparison comprising further equations:
- ##EQU69## where yi is a neighboring decoded sample value, qi is an encoded sample value, Q is the number of quantization bits for said samples, m represents the maximum quantization code (2Q -1), LQ corresponds to the lower bound of Q bits, UQ corresponds to the upper bound of Q bits, and max and min respectively corresponds maximum and minimum functions.
-
117. The computer readable medium as set forth in claim 79, wherein the encoded sample data and neighboring decoded data is grouped according to relative direction and wherein estimating comprises executable instructions, which when executed in the processing system, cause the system to perform the steps of:
-
generating directional estimates of the lost/damaged attribute data for each direction using corresponding encoded sample data and encoded neighboring data; weighting the directional estimates according relative amounts of correlation; and combining the weighted directional estimates to generate a combined estimate.
-
-
80. The computer readable medium as set forth in claim 79, wherein the encoded data comprises a video signal.
-
-
118. An apparatus configured to recover lost/damaged attribute data in a bitstream of encoded data comprising attribute data and encoded sample data, said apparatus comprising:
-
means for retrieving decoded neighboring data; and
meansfor estimating the lost/damaged attribute data using said encoded sample data, said decoded neighboring data, and available attribute data. - View Dependent Claims (119, 120, 121, 122)
-
119. The apparatus as set forth in claim 118, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, wherein said apparatus comprising means to estimate the lost/damaged attribute data by gathering square error data of an estimation function with regard to said lost/damaged DR and selecting a DR that minimizes said square error data of said estimation function.
-
120. The apparatus as set forth in claim 118, wherein lost/damaged attribute data comprises a minimum value (MIN) of the samples and available attribute data comprises a dynamic range (DR) of said samples, wherein said apparatus comprising means for estimating the lost/damaged attribute data by gathering square error data of an estimation function with regard to said lost/damaged MIN, and selecting an estimated MIN value (MIN'"'"') that minimizes said square error data of said estimation function.
-
121. The apparatus as set forth in claim 118, wherein said apparatus estimates said lost/damaged attribute data using said encoded sample data, said neighboring decoded data, and available attribute data, said apparatus further comprising:
-
means for detecting correlation between said encoded data and corresponding neighboring decoded data; means for selecting said neighboring decoded data for said estimation according to said correlation; and means for estimating said lost/damaged attribute data using said encoded sample data, said selected neighboring decoded data, and available attribute data.
-
-
122. The apparatus as set forth in claim 118, wherein the encoded sample data and neighboring decoded data is grouped according to relative direction and the apparatus comprises means to estimate by generating directional estimates of the lost/damaged attribute data for each direction using corresponding encoded sample data and encoded neighboring data, means for weighting the directional estimates according relative amounts of correlation, and means for combining the weighted directional estimates to generate a combined estimate.
-
119. The apparatus as set forth in claim 118, wherein lost/damaged attribute data comprises a dynamic range value (DR) of the samples and available attribute data comprises a minimum value (MIN) of said samples, wherein said apparatus comprising means to estimate the lost/damaged attribute data by gathering square error data of an estimation function with regard to said lost/damaged DR and selecting a DR that minimizes said square error data of said estimation function.
-
-
123. A method for recovering lost/damaged attribute data in a bitstream of encoded data comprising attribute data and encoded sample data, said method comprising:
-
retrieving decoded neighboring data; and estimating the lost/damaged attribute data using said encoded sample data and said decoded neighboring data.
-
-
124. An apparatus for recovering lost/damaged attribute data in a bitstream of encoded data comprising attribute data and encoded sample data, said method comprising:
-
means for retrieving decoded neighboring data; and means for estimating the lost/damaged attribute data using said encoded sample data and said decoded neighboring data.
-
-
125. A computer readable medium containing executable instructions which, when executed in a processing system, cause the system to recover lost/damaged attribute data in a bitstream of encoded data comprising attribute data and encoded sample data, the instructions causing the system to perform:
-
retrieving decoded neighboring data; and estimating the lost/damaged attribute data using said encoded sample data and said decoded neighboring data.
-
Specification
- Resources
Thank you for your request. You will receive a custom alert email when the Litigation Campaign Assessment is available.
×
-
Current AssigneeSony Corporation (Sony Group Corp.), Sony Electronics Inc. (Sony Group Corp.)
-
Original AssigneeSony Corporation (Sony Group Corp.), Sony Electronics Inc. (Sony Group Corp.)
-
InventorsKondo, Tetsujiro, Carrig, James J., Fujimori, Yasuhiro, Ghosal, Sugata
-
Primary Examiner(s)De Cady, Albert
-
Assistant Examiner(s)Chase, Shelly A
-
Application NumberUS09/110,803Time in Patent Office897 DaysField of Search348/421, 348/414, 348/607, 348/415, 341/94, 341/67, 714/746, 714/799US Class Current714/746CPC Class CodesH04N 19/59 involving spatial sub-sampl...H04N 19/85 using pre-processing or pos...H04N 19/88 involving rearrangement of ...H04N 19/89 involving methods or arrang...H04N 19/895 in combination with error c...H04N 19/98 Adaptive-dynamic-range codi...H04N 21/426 Internal components of the ...H04N 5/21 Circuitry for suppressing o...