Digital video protection for authenticity verification

• US 20100177891A1
• Filed: 04/11/2002
• Published: 07/15/2010
• Est. Priority Date: 04/11/2001
• Status: Active Grant

• Current Assignee: Qognify Ltd. (PSBU Midco SARL)
• Original Assignee: Nice Systems Limited (Nice Ltd)
• Inventors: Dvir, Igal, Bar-Sella, Ran, Keidar, Oren
• Granted Patent: US 7,933,407 B2
• Time in Patent Office: Days
• US Class Current: 380/200
• CPC Class Codes: G06F 21/60 Protecting data
• CPC Class Codes: G06F 21/602 Providing cryptographic facilities or services
• CPC Class Codes: G06F 21/64 Protecting data integrity, e.g. using checksums, certificates or signatures
• CPC Class Codes: G06T 1/0021 Image watermarking
• CPC Class Codes: H04N 19/467 characterised by the embedded information being invisible, e.g. watermarking
• CPC Class Codes: H04N 2005/91335 the copy protection signal being a watermark
• CPC Class Codes: H04N 2005/91364 the video signal being scrambled
• CPC Class Codes: H04N 5/913 for scrambling ; for copy protection
• CPC Class Codes: H04N 9/8042 involving data reduction

Abstract

A method and apparatus for verifying the authenticity and integrity of an ordered sequence of digital video frames, without having access to the original recording, by embedding therein a respective series of digital signatures based on a secret key, or keys, and on the video content of respective frames. Signatures are camouflaged by embedding in transform coefficients of a transformed representation of the video data in parts of the frame corresponding to motion. If there is sufficient motion to contain all of the signature bits, a supplementary technique embeds in high-texture areas of a frame. A final fall-back is to embed in a pre-defined default zone. A method of predicting when supplementary embedding is needed enables the process to be applied in a single pass allowing real-time operation. Verification is done during decoding by comparing, for identity, embedded signatures with signatures calculated anew using the method employed to embed.

Claims

1. A method for verifying an authenticity and integrity of an ordered sequence of digital video frames, the video frames including respective data, the digital frames including Inter frames and Intra frames both including segments, the segments of the Inter frames including Inter segments and Intra segments, the segments of the Intra frames including Intra segments, the method comprising the steps of:

a) providing at least one secret key; and

b) protecting the sequence by embedding a respective calculated digital signature in at least one video frame to produce a protected sequence;

i) each said calculated digital signature being dependent on the data of another video frame; and

ii) each said calculated digital signature being generated using one of said at least one secret key.

2. The method of claim 1 wherein each said at least one signature is dependent on recording environment parameters.

3. The method of claim 1 wherein each said at least one signature is embedded in a respective video frame according to an embedding pattern.

4. The method of claim 1 wherein each said at least one signature is embedded in a respective video frame according to embedding criteria.

5. The method of claim 1 wherein a respective said signature is embedded in each digital video frame.

6. The method of claim 5 wherein each said signature is dependent on an immediately preceding video frame.

7. The method of claim 1 wherein the method is performed in a single pass.

8. The method of claim 1 further including the step of:

c) compressing the video sequence into a bit stream according to a video compression standard.

9. The method of claim 8 wherein said video compression standard is a transform-based video compression standard.

10. The method of claim 9 wherein said transform-based video compression standard is a block-based video compression standard.

11. The method of claim 10 wherein said block-based video compression standard is a DCT-based video compression standard.

12. The method of claim 11 wherein said DCT-based video compression standard is selected from the group consisting of:

an H.261 video compression standard, an H.263 video compression standard, an H.263+ video compression standard, an H.263L video compression standard, an MPEG1 video compression standard, an MPEG2 video compression standard, and an MPEG4 video compression standard.

13. The method of claim 12 wherein a syntax of said video compression standard includes a frame layer.

14. The method of claim 1 wherein said digital signature is generated using an encryption algorithm.

15. The method of claim 14 wherein said encryption algorithm is a Data Encryption Standard algorithm.

16. The method of claim 14 wherein said generation is done using a full cipher block chaining mode of said Data Encryption Standard.

17. The method of claim 1 further comprising the steps of

c) dividing each segment into at least one macroblock, each said macroblock having an associated plurality of compression parameters, said compression parameters including:

i) a transform coefficient;

ii) a motion vector; and

iii) a quantization level; and

d) performing a transform on at least one respective frame thereby providing, for said respective frame, a plurality of transform coefficients wherein a respective said embedded digital signature is embedded.

18. The method of claim 17 further comprising the step of:

e) quantizing said coefficients.

19. The method of claim 18 wherein only a single bit of said embedded digital signature is embedded in each said transform coefficient.

20. The method of claim 19 wherein said transform is a Discrete Cosine Transform.

21. The method of claim 17 wherein each said transform coefficient is chosen according to an embedding pattern.

22. The method of claim 19 wherein said only a single bit is embedded in a least-significant bit of said transform coefficient.

23. The method of claim 21 wherein said embedding pattern includes a plurality of pointers to said transform coefficients.

24. The method of claim 21 wherein a respective calculated digital signal is embedded in each of a plurality of video frames according to one of a plurality of types of said embedding patterns.

25. The method of claim 24 wherein said plurality of types of said embedding patterns includes an Inter embedding pattern.

26. The method of claim 25 wherein, for each said video frame wherein a respective calculated digital signal is embedded, said Inter embedding pattern is derived from said transform coefficients of another Inter frame by selecting therefrom a plurality of said transform coefficients of highest absolute value.

27. The method of claim 26 wherein said plurality is a number greater than or equal to a number of bits of said digital signature.

28. The method of claim 26 wherein a bit of said digital signature is embedded in a transform coefficient of a macroblock pointed to by said embedding pattern only if at least one embedding criterion is satisfied.

29. The method of claim 28 wherein said at least one embedding criterion includes that an absolute value of said associated transform coefficient exceeds a predetermined value.

30. The method of claim 28 wherein said at least one embedding criterion includes that an absolute value of said associated motion vector exceeds a predetermined value.

31. The method of claim 28 wherein said at least one embedding criterion includes that said associated quantization level be at most as great as a predetermined value.

32. The method of claim 25 wherein, if said frame is an Inter frame, then in an Intra macro-block in a said Inter frame, a said transform coefficient in a said Intra macroblock pointed to by said Inter embedding pattern is embedded only if an absolute value of said associated transform coefficient exceeds a predetermined value.

33. The method of claim 26 wherein said other Inter frame is an immediately preceding Inter frame.

34. The method of claim 24 wherein said plurality of types of said embedding patterns includes a default embedding pattern including a pre-defined default embedding zone.

35. The method of claim 34 wherein a said transform coefficient is embedded if pointed to by a pointer of said default embedding pattern.

36. The method of claim 34 wherein said plurality of pointers is a number greater than or equal to a number of bits of said digital signature and wherein each said pointer points to a respective defined location in said default embedding zone.

37. The method of claim 36 wherein said default embedding zone includes a series of locations located at an end of the digital video frame according to a raster scanning order.

38. The method of claim 24 wherein said plurality of types of said embedding patterns includes an Intra embedding pattern.

39. The method of claim 38 wherein, for each said video frame wherein a respective calculated digital signal is embedded, said Intra embedding pattern is derived from said transform coefficients of another Intra frame by selecting a plurality of said pointers to highest-textured said macroblocks in said other Intra frame.

40. The method of claim 39 wherein said another Intra frame is a nearest preceding Intra frame.

41. The method of claim 39 further comprising the step of:

f) determining said highest-textured macroblocks by;

i) calculating a pixel luminance variance of each said macroblock of said other frame; and

ii) selecting said macroblocks having said plurality of highest said variances.

42. The method of claim 17 wherein, in a said Intra frame, a said transform coefficient in a said Intra macroblock is embedded only if an absolute value of said associated transform coefficient exceeds a predetermined value.

43. The method of claim 1 further comprising the steps of:

f) dividing a bit stream of the digital video frame into a plurality of consecutive blocks of bits, each said block having an equal number of bits;

g) dividing each said block into a plurality of consecutive words, each said word having an equal number of bits; and

h) i) XOR-ing said words of a first said block to provide a first block word;

ii) encrypting said first block word with said secret key, thereby obtaining an encrypted block word;

iii) XOR-ing said first encrypted block word with said words of a second said block thereby obtaining a second encrypted block word; and

vii) repeating steps (ii) and (iii) until said plurality of consecutive words have been exhausted, thereby obtaining said digital signature.

44. The method of claim 43 further including the step of padding said bit stream to enable said blocks to have equal size.

45. The method of claim 44 wherein said padding is done with zeroes.

46. The method of claim 44 further including the step of padding said blocks of bits to enable said words to have equal size.

47. The method of claim 46 wherein said padding is done with zeroes.

48. The method of claim 43 wherein said secret key includes selected data.

49. The method of claim 48 wherein said selected data includes identifying data.

50. The method of claim 49 wherein said identifying data is selected from the group consisting of information relating to a time of recording, a date of recording, and a recording channel number.

51. The method of claim 1 further comprising the step of

i) confirming an authenticity and integrity of said protected sequence.

52. The method of claim 51 wherein said confirming includes

i) extracting said embedded digital signature from each at least one frame of said protected sequence, thereby obtaining a frame signature;

ii) calculating from each in at least one frame of said protected sequence a new digital signature, using said secret key, thereby obtaining a stream signature; and

iii) inferring authenticity and integrity by comparing said extracted digital signatures with respective said new calculated signatures for identity.

53. The method of claim 6, wherein said new digital signature depends on data of another video frame.

54. The method of claim 52 wherein said frame signature in each frame of said protected sequence is extracted from a least-significant bit of said transform coefficients of claim 17 pointed to by said embedding patterns of claim 21 only if at least one criterion selected from a group consisting of:

a) that an absolute value of a said associated transform coefficient exceeds a predetermined value;

b) that an absolute value of a said associated motion vector exceeds a predetermined value; and

c) that said associated quantization level be at most as great as a predetermined value;

is satisfied.

55. The method of claim 52 wherein said stream signature is generated using an encryption algorithm.

56. The method of claim 52 wherein said authenticity and integrity are deemed confirmed if said frame signature is identical to said stream signature.

57. The method of claim 1 further comprising the step of:

j) adding a dummy, final frame to the ordered sequence of frames for containing an embedded signature to verify an original final frame of the sequence.

58. In an ordered sequence of digital video frames, the sequence being compressed according to a video compression standard, the sequence including at least one Intra frame and the frames including segments, a method of predicting a plurality of highest-textured segments in a frame comprising the steps of:

a) Counting non-zero transform coefficients in each segment of one of the at least one Intra frame; and

b) Selecting a plurality of segments having highest said counts of said transform coefficients.

59. The method of claim 58 wherein the segments are macroblocks.

60. The method of claim 58 wherein the video compression standard is a transform-based video compression standard.

61. The method of claim 60 wherein said transform-based video compression standard is a block-based video compression standard.

62. The method of claim 60 wherein said block-based video compression standard is a DCT-based video compression standard.

63. The method of claim 62 wherein said DCT-based video compression standard is selected from the group consisting of:

an H.261 video compression standard, an H.263 video compression standard, an H.263+ video compression standard, an H.263L video compression standard, an MPEG1 video compression standard, an MPEG2 video compression standard, an MPEG4 video compression standard.

64. The method of claim 58 wherein said one Intra frame is a nearest preceding Intra frame.

65. A method of protecting an authenticity and integrity of an ordered sequence of digital video frames, comprising the steps of:

a) compressing the frames according to a transform-based video compression standard; and

b) for at least one frame;

(i) embedding bits of a digital signature in respective transform coefficients of said at least one frame, and

(ii) prior to said embedding, predicting a number of mis-embeddings of said bits.

66. The method of claim 65 further comprising the steps of:

c) counting a number of non-effective embeddings in another frame; and

d) deriving the number of mis-embeddings from said number of non-effective embeddings.

67. The method of claim 66 wherein said deriving is effected according to:

M_i=max[min{(N_j−

κ

), S}, 0]

wherein M_i is the predicted number of mis-embeddings in the frame, N_j is said counted number of non-effective embeddings in the other frame, κ

is a non negative constant at most as great as a number of locations pointed to by the embedding pattern, and S is a number of bits in the digital signature.

68. The method of claim 66 wherein said other frame is an immediately preceding frame.

69. An apparatus for embedding digital signatures into an input sequence of digital video frames comprising:

a) a video encoder for compressing and encoding the input sequence to produce a compressed and encoded sequence;

b) a digital-signature-generation unit for generating a digital signature for embedding into the frames of said compressed and encoded sequence;

c) an embedding-pattern-generation unit for generating an embedding pattern; and

d) a signature-embedding unit for embedding said digital signature into the frames of said compressed and encoded sequence according to said embedding pattern.

70. The apparatus of claim 69 further comprising:

e) a memory for storing;

i) said digital signature; and

ii) said embedding pattern.

71. The apparatus of claim 69 wherein said video encoder is a transform-based video encoder.

72. The apparatus of claim 71 wherein said transform-based video encoder is a block-based video encoder.

73. The apparatus of claim 72 wherein said block-based video encoder is a DCT-based video encoder.

74. The apparatus of claim 73 wherein said DCT-based video encoder is selected from the group consisting of:

an H.261 video encoder, an H.263 video encoder, an H.263+ video encoder, an H.263L video encoder, an MPEG1 video encoder, an MPEG2 video encoder, and an MPEG4 video encoder.

75. The apparatus of claim 74 wherein:

a) said signature-embedding unit receives an input from a quantizer module of said video encoder and outputs a signal to;

i) said inverse quantizer module of said video encoder;

ii) said embedding-pattern-generation unit; and

iii) an entropy coding module of said video encoder;

b) said embedding-pattern-generation unit outputs said embedding pattern to an input of said memory for storing said embedding pattern;

c) said digital-signature-generation unit receives;

i) an output of said entropy coding module of said video encoder; and

ii) a secret key;

and outputs said digital signature to another input of said memory for storing said digital signature; and

d) said memory for storing outputs said embedding pattern and said digital signature to respective inputs of said signature embedding unit.

76. An apparatus for authenticating and verifying respective embedded digital signatures in an input sequence of digital frames comprising:

a) a video decoder for decompressing and decoding the input sequence, thereby providing a decompressed and decoded sequence;

b) a stream-signature generation unit for calculating a digital signature for each frame of said decompressed and decoded sequence;

c) an embedding pattern generating unit for generating a respective embedding pattern for each frame of said decompressed and decoded sequence to point to a location, in said each frame, of the respective embedded digital signature;

d) a frame signature extracting unit for extracting the respective embedded digital signature from each frame of said decompressed and decoded signature in accordance with said respective embedding pattern; and

e) an authenticity-verification unit for determining an authenticity and veracity of said decompressed and decoded sequence.

77. The apparatus of claim 76 further comprising:

f) at least one memory for;

(i) storing said calculated digital signature;

(ii) storing said extracted embedded digital signature; and

(iii) storing said generated embedding pattern.

78. The apparatus of claim 76 wherein said video decoder is a transform-based augmented video decoder.

79. The apparatus of claim 78 wherein said transform-based augmented video decoder is a block-based augmented video decoder.

80. The apparatus of claim 79 wherein said block-based augmented video decoder is a DCT-based augmented video decoder.

81. The apparatus of claim 80 wherein said DCT-based augmented video decoder is selected from the group consisting of:

an H.261 augmented video decoder, an H.263 augmented video decoder, an H.263+ augmented video decoder, an H.263L augmented video decoder, an MPEG1 augmented video decoder, an MPEG2 augmented video decoder, and an MPEG4 augmented video decoder.

82. The apparatus of claim 81 wherein:

a) said frame signature extraction unit receives an input from an entropy decoding module of said video decoder and outputs a signal to an inverse quantizer module of said video decoder;

b) said embedding pattern generation unit receives an input from said entropy decoding module of said video decoder and outputs said generated embedding pattern to said memory for storing said generated embedding pattern;

c) said memory for storing said generated embedding pattern outputs said generated embedding pattern to said frame signature extraction unit;

d) said stream signature generation unit receives;

i) a same input as said entropy decoding module; and

ii) a secret key;

and outputs said calculated digital signature to said memory for storing said calculated digital signature;

e) said memory for storing said calculated digital outputs said calculated digital signature to said authenticity-verification unit; and

f) said authenticity-verification unit also receives a said extracted embedded digital signature from said frame-signature extraction unit and outputs an authenticity decision.