Digital watermark embedding apparatus, digital watermark detecting apparatus, digital watermark embedding method, digital watermark detecting method and computer program product

• US 7,826,637 B2
• Filed: 01/23/2006
• Issued: 11/02/2010
• Est. Priority Date: 10/31/2000
• Status: Expired due to Fees

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Muratani, Hirofumi
• Primary Examiner(s): STREGE, JOHN B
• Application Number:
  US11/336,832
• Publication Number:
  US 20060120561A1
• Time in Patent Office: 1,744 Days
• Field of Search: None
• US Class Current: 382/100
• CPC Class Codes: G06T 1/0064 Geometric transfor invariant watermarking, e.g. affine transform invariant
• CPC Class Codes: G06T 1/0085 Time domain based watermarking, e.g. watermarks spread over several images
• CPC Class Codes: G06T 2201/0061 Embedding of the watermark in each block of the image, e.g. segmented watermarking
• CPC Class Codes: G10L 19/018 Audio watermarking, i.e. embedding inaudible data in the audio signal
• CPC Class Codes: H04N 1/32144 embedded in the image data, i.e. enclosed or integrated in the image, e.g. watermark, super-imposed logo or stamp
• CPC Class Codes: H04N 1/32267 combined with processing of the image
• CPC Class Codes: H04N 1/32352 Controlling detectability or arrangements to facilitate detection or retrieval of the embedded information, e.g. using markers for decoding, detection or retrieval operations H04N1/32149 and subgroups take precedence
• CPC Class Codes: H04N 2201/3233 of authentication information, e.g. digital signature, watermark
• CPC Class Codes: H04N 2201/327 which are undetectable to the naked eye, e.g. embedded codes

Abstract

A digital watermark embedding apparatus for embedding watermark information into a digital content, obtains a predetermined topological invariant corresponding to the watermark information to be embedded into the digital content, and sets the predetermined topological invariant to the digital content by modifying the digital content into which the watermark information is to be embedded.

Claims

1. A watermark embedding apparatus, comprising:

a processor comprising;

a homotopy class calculating unit configured to calculate a homotopy class of an input image to be watermarked;

a watermark information class transforming unit configured to calculate a homotopy class of watermark information; and

a homotopy class changing unit configured to change at least a portion of the input image based on a topological invariant, wherein the topological invariant is maintained when a local distortion which is a homeomorphism is applied to a topological space formed of a set of points of the input image, the local distortion including a Stirmark attack, the topological invariant corresponding to watermark information to be embedded into the input image, and the local distortion being expressed by functions, each of the functions depending on a position of the points;

wherein the homotopy class changing unit is further configured to change said at least said portion of the input image so that the homotopy class of the input image corresponds to the homotopy class of the watermark information.

2. The apparatus according to claim 1, wherein

the topological invariant is a homotopy invariant.

3. The apparatus according to claim 2, wherein

said changing unit is further configured to change a predetermined part of a bit string corresponding to a spatial or time-basis position of the input image to a target value determined by the position of the bit string and the topological invariant or a target restorable value.

4. The apparatus according to claim 3, wherein

said changing unit is further configured to generate a mapping to a target topological space defined by the predetermined part of said bit string from a base topological space defined by the spatial or time-basis position, and obtain the target value corresponding to the position based on a function for applying the topological invariant.

5. The apparatus according to claim 4, wherein

the function has the topological invariant as a parameter.

6. The apparatus according to claim 3, wherein

the target restorable value is a value obtained by scaling the target value, and

the changing unit is further configured to select a scale power for scaling the target value so as to decrease the influence on the input image for the target value corresponding to the position.

7. The apparatus according to claim 4, wherein

the changing unit is further configured to evaluate an influence on the input image if the topological invariant is set to the input image using a plurality of different functions, and to use a function evaluated to minimize the influence on the input image as the function for applying the topological invariant.

8. The apparatus according to claim 2, wherein

said changing unit is further configured to set a value, allowing a representative value representing a divided region which is obtained by dividing the input image in a spatial manner or a time-basis manner and obtained from a predetermined part of one or a plurality of bit strings corresponding to a spatial or time-basis position of the divided region to become a target value determined by the spatial or time-basis position of the divided region and the topological invariant or a target restorable value, to the predetermined part of the one or the plurality of bit strings.

9. The apparatus according to claim 8, wherein

the bit string corresponding to the plurality of positions enables configuring a content of one divided region, and

the representative value is an average value of values of the predetermined parts of the plurality of bit strings.

10. The apparatus according to claim 2, wherein

the input image into which the watermark information is to be embedded is a content dividable into a plurality sub-contents, and the topological invariant corresponding to the watermark information is set to the plurality of sub-contents.

11. The apparatus according to claim 2, wherein

the input image into which the watermark information is to be embedded is a content dividable into a plurality sub-contents, and a value according to a pseudo-random number generated based on the watermark information is set, as the topological invariant, to the plurality of sub-contents.

12. The apparatus according to claim 2, wherein the topological invariant comprises watermark information to be embedded into the input image such that the watermark is not lost.

13. A watermark detecting apparatus, comprising:

a processor comprising;

a homotopy class calculating unit configured to calculate a homotopy class of a watermarked image; and

a detector configured to detect watermark information by obtaining a topological invariant, wherein the watermarked image comprises points, the topological invariant is maintained when a local distortion which is a homeomorphism is applied to a topological space formed of a set of points of the watermarked image, the local distortion including a Stirmark attack, the local distortion being expressed by functions, each of the functions depending on a position of the points;

wherein the detector is further configured to obtain watermark information corresponding to the calculated homotopy class.

14. The apparatus according to claim 13, wherein

the topological invariant is a homotopy invariant.

15. The apparatus according to claim 14, wherein

said detector is further configured to obtain a representative value representing a divided region obtained by dividing the watermarked image in a spatial manner or a time-basis manner based on a value read from a predetermined part of one or a plurality of bit strings corresponding to a spatial or time-basis position of the divided region, and obtain the topological invariant based on the representative value representing the divided region.

16. The apparatus according to claim 15, wherein

a plurality of positions are provided, a watermarked image of one divided region is constituted by the bit string corresponding to the plurality of positions, and the representative value is an average value of values of the predetermined part of the plurality of bit strings.

17. The apparatus according to claim 15, wherein

said detector is further configured to obtain a target value normalizing the representative value of the divided region, and perform a predetermined summing arithmetic operation for the target value and a difference in target values, thereby obtaining the topological invariant.

18. The apparatus according to claim 14, wherein

the watermarked image into which the watermark information is to be embedded is a content dividable into a plurality of sub-contents, topological invariants are detected from the plurality of sub-contents, respectively, and the watermark information is obtained based on the obtained topological invariants.

19. The apparatus according to claim 14, wherein

the watermarked image is image data of a still image or moving image frames, and

the part is a pixel value vector of the image data.

20. The apparatus according to claim 14, wherein

the watermarked image is moving image data, and

the part is a pixel value vector of the moving image data.

21. The apparatus according to claim 14, wherein

the watermarked image is voice data, and

the part is an amplitude value vector of the voice data.

22. The apparatus according to claim 13, wherein the topological invariant comprises watermark information to be embedded into the watermarked image such that the watermark is not lost.

23. A non-transitory computer readable medium having computer executable instructions, which when executed by a computer, cause the computer to perform a method comprising:

calculating a homotopy class of a watermarked image

detecting watermark information of the watermarked image by obtaining a topological invariant, wherein the watermarked image comprises points, the topological invariant is maintained when a local distortion which is a homeomorphism is applied to a topological space formed of a set of points of the watermarked image, the local distortion including a Stirmark attack, the local distortion being expressed by functions, each of the functions depending on a position of the points; and

obtaining watermark information corresponding to the calculated homotopy class.

24. The medium according to claim 23, wherein the topological invariant comprises watermark information to be embedded into the watermarked image such that the watermark is not lost.

25. The medium according to claim 23, wherein the toplogical invariant is a homotopy invariant.

26. A watermark embedding apparatus, comprising:

a processor comprising;

a watermark information class transforming unit configured to calculate a homotopy class of watermark information; and

a homotopy class changing unit configured to generate an embedding function corresponding to an element of the homotopy class of the watermark information and to change at least a portion of an input image based on a topological invariant, wherein the topological invariant is maintained when a local distortion which is a homeomorphism is applied to a topological space formed of a set of points of the input image, the local distortion including a Stirmark attack, the topological invariant corresponding to the embedding function and the local distortion being expressed by functions, each of the functions depending on a position of the points;

wherein the homotopy class changing unit is further configured to change said at least said portion of the input image so that the homotopy class of the input image corresponds to the homotopy class of the watermark information.

27. A watermark embedding apparatus, comprising:

a processor comprising;

a frame dividing unit configured to divide a moving image into frames;

a watermark information class transforming unit configured to calculate a homotopy class of watermark information;

a homotopy class changing unit configured to change at least a portion of each of the frames based on a topological invariant, wherein the topological invariant is maintained when a local distortion which is a homeomorphism is applied to a topological space formed of a set of points of a frame, the local distortion including a Stirmark attack, the topological invariant corresponding to watermark information to be embedded into moving image data, and the local distortion being expressed by functions, each of the functions depending on a position of the points; and

a frame composing unit configured to compose the frames having a portion which is changed into the moving image data;

wherein the homotopy class changing unit is further configured to change said at least said portion of each of the frames so that the homotopy class of the frames corresponds to the homotopy class of the watermark information.

28. A watermark detecting apparatus, comprising:

a processor comprising;

a frame dividing unit configured to divide a moving image into frames; and

a detector configured to detect watermark information by obtaining a topological invariant, wherein the frame comprises points, the topological invariant is maintained when a local distortion which is a homeomorphism is applied to a topological space formed of a set of points of the frame, the local distortion including a Stirmark attack, the local distortion being expressed by functions, each of the functions depending on a position of the points;

wherein the detector is further configured to obtain watermark information corresponding to a calculated homotopy class.

29. A non-transitory computer readable medium having computer executable instructions, which when executed by a computer, cause the computer to perform a method comprising:

dividing a watermarked moving image into frames;

detecting watermark information of the watermarked image by obtaining a topological invariant, wherein the frame comprises points, the topological invariant is maintained when a local distortion which is a homeomorphism is applied to a topological space formed of a set of points of the frame, the local distortion including a Stirmark attack, the local distortion being expressed by functions, each of the functions depending on a position of the points; and

obtaining watermark information corresponding to a calculated homotopy class.