A fingerprinting scheme for piracy tracking and prosecution of digital works

Abstract

This invention relates to a fingerprinting scheme for tracking and prosecuting piracy of digital works, belonging to the field of digital copyright protection technology. It includes the process of generating fingerprint-containing digital products and the fingerprint identification process after discovering illegal copies. The scheme does not require a trusted third party or blockchain technology support. Through a two-way constraint mechanism between publishers and users, it achieves accurate tracking and prosecution of piracy. This invention requires neither blockchain technology nor a trusted third party, yet it can accurately determine the subject of illegal copying—whether it is a user who committed illegal copying or a publisher who copied and then framed others—thus avoiding affecting the practicality of the scheme. The execution process is simple and efficient, avoiding the consumption of large amounts of computing, storage, and communication resources.

Description

Technical Field

[0001] This invention relates to a fingerprinting scheme for tracking and prosecuting piracy of digital works, belonging to the field of digital copyright protection technology. Background Technology

[0002] In the digital age, the copying and dissemination of digital works (such as music, videos, and e-books) has become extremely convenient, providing opportunities for piracy. To protect the intellectual property rights of digital works, the industry widely adopts digital watermarking and digital fingerprinting technologies. These technologies allow sellers to trace the source of illegal copies by extracting embedded digital watermarks after discovering them, and then use digital fingerprints to prosecute the illegal copyists.

[0003] However, most existing digital watermarking technologies and digital fingerprinting solutions often require the involvement of a trusted third party (such as a copyright center) or rely on blockchain technology for implementation. This not only consumes significant computing, storage, and communication resources but may also affect the practicality of the solution due to fluctuations in the trustworthiness of the third party or the technical barriers of blockchain. Therefore, this invention provides a fingerprinting solution for tracking and prosecuting piracy of digital works. Summary of the Invention

[0004] In view of this, the present invention provides a fingerprinting scheme for tracking and accusing piracy of digital works. It does not require blockchain technology or trusted third-party participation, but can accurately determine the subject of illegal copying behavior, whether it is a user who has committed illegal copying or the publisher who copied it and then framed others, so as not to affect the practicality of the scheme. The execution process is simple and efficient, avoiding the consumption of a lot of computing, storage and communication resources.

[0005] This invention provides a fingerprint scheme for tracking and prosecuting piracy of digital works. The proposed technical solution includes a process for generating fingerprint-containing digital products and a fingerprint identification process after discovering illegal copies. The process of generating the fingerprint-integrated digital product includes: S1: Generate a product for sale containing a tracking fingerprint: In response to a user's purchase request, the publisher embeds a unique tracking fingerprint TF into the original digital product X, generating a product for sale XU. S2: Inform users of product sampling information: The publisher informs users of the number of parameters L of the product XU to be sold; S3: User generates and encrypts secret vector: The user generates a secret vector SV with 2L elements, each element having a value evenly distributed in the interval (1-ε, 1+ε). After being encoded with 256 bits, the user encrypts the SVD using their own public key kd and sends it to the issuer. S4: Publisher verifies and determines the charge fingerprint: The publisher randomly divides the SVD into D1 and D2. After verifying D1, D2 is determined as the charge fingerprint. S5: The issuer scrambles and encrypts the accusation fingerprint D2: The issuer uses a secret scrambling algorithm ψ to scramble and encrypt D2, resulting in PT; S6: Generate the final encrypted product: The issuer combines XU, PT and the user's public key kd to generate the encrypted product SUV and sends it to the user; S7: Publisher sends encrypted product to user: The publisher sends the Vector SUV as an encrypted product to the user; S8: User decrypts to obtain product containing fingerprints: The user uses the private key ke to decrypt SUV and obtains the final digital work Y containing both the tracking fingerprint TF and the accusation fingerprint D2; The fingerprint identification process following the discovery of illegally copied fingerprints includes: Step 1: User generates and encrypts random data: The user generates and encrypts random data T, calculates its hash value SH, and sends it to the publisher for archiving; Step II: Publisher Group Hash Value: The publisher randomly groups the SH and informs the user of the grouping results; Step III: User generates test random vector set: The user generates test random vector set ST according to the grouping; Step IV: User generates test signals and submits them for authentication: The user generates K test signals using ST and the accusation fingerprint D2, including (K-1) fake fingerprint test signals and 1 real fingerprint test signal; Step V: User discloses partial random vector information: The user discloses the 1st to 2K-2th random vectors generated in Step III, and the 2K-1th random vector st. 2K-1 The encrypted value TD 2K-1 ; Step VI: The publisher verifies the legitimacy of the test signal and determines whether the user has committed illegal copying based on whether the signal accurately points to a real fingerprint.

[0006] Furthermore, in step S3, the user encodes each element in SV using 256 bits. The user must keep SV strictly confidential and then encrypts each element in SV one by one using their own public key kd. The algorithm formula is: SVD = E kd (SV).

[0007] Furthermore, in step S5, the scrambling algorithm ψ and its parameters are selected by the publisher and kept secret. Its function is to shuffle the order of D2 to ensure that the user cannot delete D2 from the received digital product. The algorithm formula is: PT=ψ(D2).

[0008] Furthermore, in step S6, let SV2 be the part of SV that corresponds to D2. The issuer will encode each element in XU with 256 bits, encrypt each value with the user's public key kd, and finally multiply it with the corresponding position of ST to generate the vector SUV. The algorithm formula is as follows: SUV = E kd (XU)×PT=E kd (XU)×ψ(D2)=E kd (XU)×ψ(E kd (SV2))=E kd (XU×ψ(SV2)).

[0009] Furthermore, in step S7, the part of SV that corresponds to D2 is denoted as SV2, and the algorithm formula is: Y=D ke (SUV)=D ke (E kd (XU×ψ(SV2)))=XU*ψ(SV2).

[0010] Furthermore, in step I, the user encodes each element in T using 256 bits, and then encrypts each encoded element using their own public key to obtain TD. The algorithm formula is: td i =E kd (t i ), i=1,2, (2K-1)×L, then, the user calculates the hash value SH of each element in TD, using the following formula: SH={hash (E kd (ti)), i=1,2, ,(2K-1)×L}.

[0011] Furthermore, in step III, the algorithm formula for constructing the vector group ST is: ST = {st i ,i=1,L2K-1}, where each st i Both are vectors of size L, whose elements are all taken from T, and all elements are unique.

[0012] Furthermore, in step IV, the algorithm formula for generating the fake fingerprint test signal is: t i =st 2i-1 ⊕st 2i i=1,2, K-1, t K =st 2K-1 ⊕D2; where ⊕ is the XOR operation.

[0013] Furthermore, in step V, TD 2K-1 It uses the user's public key for encryption, and the algorithm formula is: TD 2K-1=E kd (st 2K-1 ).

[0014] The beneficial effects of this invention are: it requires neither blockchain technology nor the participation of a trusted third party, yet it can accurately determine the subject of illegal copying, whether it is a user who illegally copied the work or a publisher who copied it and then blamed it on others, thus avoiding affecting the practicality of the solution. The execution process is simple and efficient, avoiding the consumption of a large amount of computing, storage, and communication resources. Moreover, its security has been verified through multiple layers, providing a practical and effective solution for tracking and accusing piracy of digital works, and it can serve as a sales agreement for graphic and image-based digital works. Detailed Implementation

[0015] The preferred embodiments of the present invention will now be described in detail.

[0016] This invention provides a fingerprinting scheme for tracking and prosecuting piracy of digital works, including the generation process of fingerprint-containing digital products and the fingerprint identification process after illegal copies are discovered. The scheme does not require the support of a trusted third party or blockchain technology, and achieves accurate tracking and prosecution of piracy through a two-way constraint mechanism between publishers and users.

[0017] As an optional embodiment, the process of generating a fingerprint-integrated digital product includes: S1: Generating a Product for Sale with a Tracking Fingerprint: In response to a user's purchase request, the publisher embeds a unique tracking fingerprint TF into the original digital product X, generating a product for sale XU. Specifically, when a user requests to purchase a multimedia digital product, the publisher first embeds a digital fingerprint TF for tracking into the original product X, generating a product for sale XU. It should be noted that the tracking fingerprint TF is unique for each product purchased by a user. This fingerprint is only used to initially locate the source of the illegal copy and cannot be directly used as legal evidence to accuse the illegal copyist. S2: Informing the user of product sampling information: The publisher informs the user of the number of parameters L of the product XU to be sold. This parameter L provides the basis for the user to subsequently generate an accusation fingerprint, ensuring that the fingerprint generated by the user matches the product characteristics. Specifically, the publisher informs the user who makes the purchase request of the number of parameters L of the digital work XU. For example, for most digital images, an 8*8 pixel image block can be used as a parameter. If each pixel is represented by 8 bits, then the length of each parameter is 256 bits. This information is the basic parameter for the user to subsequently generate an "accusation fingerprint," ensuring that the fingerprint generated by the user matches the product characteristics. S3: User generates and encrypts the secret vector: The user generates a secret vector SV with 2L elements, each element having a value evenly distributed in the interval (1-ε, 1+ε) (ε is a small value to ensure that the value of each element is close to 1). The user encodes each element in SV using 256 bits. The user must keep SV strictly confidential and then encrypt each element in SV one by one using their own public key kd. The algorithm formula is: SVD = E kd (SV), after encryption, the user sends the encrypted SVD to the publisher; S4: Publisher verifies and determines the accusation fingerprint: The publisher randomly divides the SVD into D1 and D2. After verifying D1, D2 is determined as the accusation fingerprint. Specifically, after receiving the SVD, the publisher randomly divides it into two equal parts, D1 and D2. Subsequently, the publisher requires the user to decrypt D1 and verify whether the decrypted data indeed follows a uniform distribution within the interval (1-ε, 1+ε). This step is to ensure that the secret vector generated by the user conforms to the rules and to prevent the user from maliciously constructing data that affects subsequent processes. If the verification passes, the publisher determines the uncracked part D2 as the accusation fingerprint and requires the user to sign D2 (the signature further ensures that D2 belongs to the user). If the verification fails, it indicates that the user has violated regulations, and the publisher has the right to refuse to sell the digital product. S5: The issuer scrambles and encrypts the accusation fingerprint D2: The issuer uses a secret scrambling algorithm ψ to scramble and encrypt D2, resulting in PT; the scrambling algorithm ψ and its parameters are selected by the issuer and kept secret from outsiders. Its function is to shuffle the order of D2 (without changing the data itself) to ensure that users cannot delete D2 from the received digital product. The algorithm formula is: PT=ψ(D2). S6: Generating the final encrypted product: The issuer combines XU, PT, and the user's public key kd to generate the encrypted product SUV and sends it to the user; SV2 is the part of SV that corresponds to D2. The issuer encodes each element in XU with 256 bits, encrypts each value using the user's public key kd, and finally multiplies it with the corresponding position in ST to generate the vector SUV. The algorithm formula is: SUV = E kd (XU)×PT=E kd (XU)×ψ(D2)=E kd (XU)×ψ(E kd (SV2))=E kd (XU×ψ(SV2)); S7: Publisher sends encrypted product to user: The publisher sends the Vector SUV as an encrypted product to the user; S8: User decrypts to obtain the product containing fingerprints: The user uses the private key ke to decrypt SUV, obtaining the final digital work Y containing both the tracking fingerprint TF and the accusation fingerprint D2; let SV2 be the part of SV that corresponds to D2, and the algorithm formula is: Y=D ke (SUV)=D ke (E kd (XU×ψ(SV2)))=XU*ψ(SV2); Because each element of SV has a value close to 1 (within the range (1-ε, 1+ε)), and SV2 is a subset of SV, each element of SV2 also has a value close to 1 (within the range (1-ε, 1+ε)). Y will be very close to XU, meaning it will not affect the user's experience with the digital work, but the work already contains the digital works of TF and D2. This process achieves two-way constraints through technical design: the publisher cannot generate a usable product Y because it does not know the user's private key; the user cannot remove the embedded accusation fingerprint D2 because it does not know the scrambling method ψ. Therefore, in the digital work that the user finally obtains, neither the tracking fingerprint nor the accusation fingerprint can be unilaterally tampered with, laying the foundation for subsequent piracy tracking and accusation. It should be noted that if the elements of D2 deviate significantly from the range (1-ε, 1+ε), the deviation between Y and XU will be large, causing the product purchased by the user to lose its usability. This is also the key reason why step 3 requires the SV element to have a value close to 1.

[0018] As an optional implementation, when a publisher discovers illegal copies, it needs to identify the entity responsible for the illegal copying through the following steps to ensure the accuracy and impartiality of the accusation: The fingerprint identification process after discovering illegal copies includes: Step 1: User generates and encrypts random data: The user generates (2K-1) × L random data Ts uniformly distributed in the interval (1-ε, 1+ε) (i.e., each element Ti in T ∈ (1-ε, 1+ε)), where L has the same meaning as before, and K is an integer used to reduce the probability of the issuer falsely accusing the user (the larger the value of K, the lower the probability of false accusation, but the greater the computational cost); then, the user encodes each element in T with 256 bits, and encrypts each encoded element one by one with their own public key to obtain TD. The algorithm formula is: td i =E kd (t i ), i=1,2, (2K-1)×L, then, the user calculates the hash value SH of each element in TD, using the following formula: SH={hash (E kd (ti)), i=1,2, Finally, the user sends the SH to the publisher; Step II: Publisher Grouping Hash Values: The publisher randomly groups the SH and informs the user of the grouping results; specifically, after receiving the SH, the publisher first records and archives it, then randomly groups the (2K-1)×L elements in the SH, selecting L elements from each group, thus forming (2K-1) groups in total. j (j=1,2,3,...,2K-1)); afterwards, the publisher will inform the user of the grouping results; Step III: User generates a test random vector group: The user generates a test random vector group ST based on the grouping; the user groups the random data T generated in Step I according to the same rules based on the publisher's grouping method, generating 2K-1 test random vectors. The algorithm formula for forming the vector group ST is: ST = {st i ,i=1,L2K-1}, where each st i Both are vectors of size L, whose elements are all taken from T, and all elements are unique. Step IV: User generates test signals and submits them for authentication: The user generates K test signals using ST and the accusation fingerprint D2, including (K-1) fake fingerprint test signals and 1 real fingerprint test signal; the algorithm formula for generating fake fingerprint test signals is: t i =st 2i-1 ⊕st 2i i=1,2, K-1, t K =st 2K-1 ⊕D2; where ⊕ is the XOR operation; specifically, after generation, the user will scramble these K test signals as a whole (without changing the values ​​of elements within the group, only shuffling the group order), and then send them to the issuer, requesting the issuer to indicate which signal contains the real fingerprint D2; At this point, if the publisher did indeed obtain D2 from the illegal copy, they can determine this by calculating the cross-correlation coefficient between D2 and each test signal: since the elements of D2 follow a uniform distribution in the interval (1-ε, 1+ε), if the test signal does not contain a real fingerprint, the cross-correlation coefficient will be close to 0; if it contains a real fingerprint D2, the cross-correlation coefficient will be significantly greater than 0. Therefore, the publisher can accurately identify the test signal containing a real fingerprint. If the publisher attempts to frame a legitimate user, since they do not have the real D2, they can only guess randomly, and the probability of correctly pointing it out is only 1 / K. When K is a large value (such as K=100), this probability will be very low, such as 1% when K=100, which can almost eliminate the possibility of framing. Step V: User discloses partial random vector information: The user discloses the 1st to 2K-2th random vectors generated in Step III, and the 2K-1th random vector st. 2K-1 The encrypted value TD 2K-1 ;TD2K-1 It uses the user's public key for encryption, and the algorithm formula is: TD 2K-1 =E kd (st 2K-1 This step is to allow the publisher to verify the legitimacy of the test signal, while avoiding the leakage of the real fingerprint information; Step VI: The publisher verifies the legality of the test signal and determines whether the user has committed illegal copying based on whether the signal accurately points to a real fingerprint. Specifically, the publisher needs to verify whether the following three conditions are met: (1) Whether the elements of the publicly available 2K-2 random vectors all follow a uniform distribution in the interval (1-ε, 1+ε); (2) By calculating st 2K-1 The hash value of each element is used to confirm whether it belongs to the originally received SH (to ensure that the data has not been tampered with); (3) the encrypted real fingerprint test signal D 2K-1 Is it related to st 2K-1 ⊕D2 is the same. If any one of the three conditions is not met, it means that the user has committed fraud during the authentication process; if all three conditions are met, and the issuer accurately points out the test signal containing the real fingerprint, it can be confirmed that the user has committed illegal copying. At the same time, because the user's original data is not publicly available, the publisher cannot deduce D2 from the test signal, thus avoiding the risk that the publisher may retain D2 for subsequent false accusations.

[0019] In summary, this solution achieves three core advantages through its "dual fingerprint" design and two-way constraint mechanism: No third-party support required: It eliminates reliance on trusted third parties or blockchain, reducing resource consumption and technical barriers; Anti-tampering and anti-framing: Publishers cannot generate accusation fingerprints independently, users cannot delete fingerprints, and the K-value setting during the authentication process significantly reduces the probability of false accusations; Simple and secure process: Every step from product generation to illegal copy authentication is supported by clear technical logic, ensuring the normal use of digital works while providing reliable evidence for piracy accusations; This solution provides an efficient, secure, and easily implementable solution for the intellectual property protection of digital works and can serve as a sales agreement for graphic and image-based digital works.

[0020] The present invention and its embodiments have been described above. This description is not restrictive, and the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and, without departing from the spirit of the invention, design similar structures and embodiments without creative effort, all such designs should fall within the protection scope of the present invention.

Claims

1. A fingerprinting scheme for tracking and prosecuting piracy of digital works, characterized in that: This includes the process of generating digital products containing fingerprints and the fingerprint identification process after discovering illegally copied products; The process of generating the fingerprint-integrated digital product includes: S1: Generate a product for sale containing a tracking fingerprint: In response to a user's purchase request, the publisher embeds a unique tracking fingerprint TF into the original digital product X, generating a product for sale XU; S2: Inform users of product sampling information: The publisher informs users of the number of parameters L of the product XU to be sold; S3: User generates and encrypts secret vector: The user generates a secret vector SV with 2L elements, each element having a value evenly distributed in the interval (1-ε, 1+ε). After being encoded with 256 bits, the user encrypts the SVD using their own public key kd and sends it to the issuer. S4: Publisher verifies and determines the charge fingerprint: The publisher randomly divides the SVD into D1 and D2. After verifying D1, D2 is determined as the charge fingerprint. S5: The issuer scrambles and encrypts the accusation fingerprint D2: The issuer uses a secret scrambling algorithm ψ to scramble and encrypt D2, resulting in PT; S6: Generate the final encrypted product: The issuer combines XU, PT and the user's public key kd to generate the encrypted product SUV and sends it to the user; S7: Publisher sends encrypted product to user: The publisher sends the Vector SUV as an encrypted product to the user; S8: User decrypts to obtain product containing fingerprints: The user uses the private key ke to decrypt SUV and obtains the final digital work Y containing both the tracking fingerprint TF and the accusation fingerprint D2; The fingerprint identification process following the discovery of illegally copied fingerprints includes: Step 1: User generates and encrypts random data: The user generates and encrypts random data T, calculates its hash value SH, and sends it to the publisher for archiving; Step II: Publisher Group Hash Value: The publisher randomly groups the SH and informs the user of the grouping results; Step III: User generates test random vector set: The user generates test random vector set ST according to the grouping; Step IV: User generates test signals and submits them for authentication: The user generates K test signals using ST and the accusation fingerprint D2, including (K-1) fake fingerprint test signals and 1 real fingerprint test signal; Step V: User discloses part of the random vector information: The user discloses the 1 to 2K-2 random vectors generated in Step III, and the encrypted value TD of the 2K-1 random vector st 2K-1 2K-1 ;​ Step VI: The publisher verifies the legitimacy of the test signal and determines whether the user has committed illegal copying based on whether the signal accurately points to a real fingerprint.

2. The fingerprint scheme for tracking and prosecuting piracy of digital works according to claim 1, characterized in that: In step S3, the user uses 256-bit to encode each element in SV, and the user needs to strictly keep SV secret, and then uses the user's public key kd to encrypt each element in SV one by one, and the algorithm formula is: SVD=E kd (SV).

3. The fingerprint scheme for tracking and prosecuting piracy of digital works according to claim 1, characterized in that: In step S5, the scrambling algorithm ψ and its parameters are selected by the publisher and kept secret. Its function is to shuffle the order of D2 to ensure that users cannot delete D2 from the received digital product. The algorithm formula is: PT=ψ(D2).

4. A fingerprint scheme for tracking and prosecuting piracy of digital works according to claim 1, characterized in that: In step S6, the part of SV corresponding to D2 is denoted as SV2. The issuer encodes each element in XU with 256 bits, encrypts each value using the user's public key kd, and finally multiplies it with the corresponding position of ST to generate the vector SUV. The algorithm formula is: SUV = E kd (XU)×PT=E kd (XU)×ψ(D2)=E kd (XU)×ψ(E kd (SV2))=E kd (XU×ψ(SV2)).

5. A fingerprint scheme for tracking and prosecuting piracy of digital works according to claim 1, characterized in that: In step S8, the part of SV that corresponds to D2 is denoted as SV2, and the algorithm formula is: Y=D ke (SUV)=D ke (E kd (XU×ψ(SV2)))=XU*ψ(SV2).

6. A fingerprint scheme for tracking and prosecuting piracy of digital works according to claim 1, characterized in that: In step I, the user encodes each element in T using 256 bits, and then encrypts each encoded element using their own public key to obtain TD. The algorithm formula is: td i =E kd (t i ), i=1,2, (2K-1)×L, then, the user calculates the hash value SH of each element in TD, using the following formula: SH={hash (E kd (ti)), i=1,2, ,(2K-1)×L}.

7. A fingerprint scheme for tracking and prosecuting piracy of digital works according to claim 1, characterized in that: In step III, the algorithm formula for constructing the vector group ST is: ST = {st i ,i=1,L2K-1}, where each st i Both are vectors of size L, whose elements are all taken from T, and all elements are unique.

8. A fingerprint scheme for tracking and prosecuting piracy of digital works according to claim 1, characterized in that: In step IV, the algorithm formula for generating the fake fingerprint test signal is: t i =st 2i-1 ⊕st 2i i=1,2, K-1, t K =st 2K-1 ⊕D2; where ⊕ is the XOR operation.

9. A fingerprint scheme for tracking and prosecuting piracy of digital works according to claim 1, characterized in that: In step V, TD 2K-1 It uses the user's public key for encryption, and the algorithm formula is: TD 2K-1 =E kd (st 2K-1 ).