High-capacity low-distortion watermarking method for digital artwork copyright protection

Abstract

The invention discloses a high-capacity low-distortion watermarking method for digital artwork copyright protection, which comprises the following steps: acquiring a digital artwork image to be protected, and embedding robust watermark information to obtain a digital artwork image containing the robust watermark information; embedding the reversible watermark into the digital artwork image containing the robust watermark information to obtain a digital artwork image containing the robust watermark information and the reversible watermark; processing the digital artwork image containing the robust watermark information and the reversible watermark and then sending the processed digital artwork image to a receiving end; and if the receiving end passes the integrity verification, carrying out watermark extraction and image restoration on the received digital artwork image containing the robust watermark information and the reversible watermark, otherwise, carrying out watermark extraction on the received digital artwork image containing the robust watermark information and the reversible watermark. Watermark embedding is carried out in combination with discrete inseparable shear wave transformation and local rapid generalized polar coordinate complex exponential transformation, and imperceptibility, robustness and effective watermark capacity are improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a high-capacity and low-distortion watermark method for copyright protection of digital artworks. Background Art

[0002] The creation, dissemination and trading of digital image artworks are becoming increasingly popular, and the copyright protection of digital image artworks has become very urgent. Digital watermarking technology is one of the effective methods for copyright protection of digital media content. It is worth noting that digital image artworks often have high requirements for visual quality, and image distortion is unacceptable for digital image artworks. In traditional watermarking technology, both robust watermarking technology and fragile / semi-fragile watermarking technology will cause certain distortion to the carrier image. Since the distortion caused by watermarking technology is not compensated, this distortion is difficult to detect by the human visual system, but it is unacceptable for images in some specific scenes, such as digital image artworks, military images, medical images, etc. The introduction of reversible watermarking technology solves the problem of carrier image distortion. After extracting the watermark information from the watermarked image, this technology can eliminate the distortion caused by watermark embedding, that is, restore the watermarked image to a watermark-free state, which has a lossless nature. Therefore, this technology is widely used in the copyright protection of high-assurance images. Although reversible watermarking technology solves the distortion problem, the robustness of this watermarking technology is weak. Even if it is slightly attacked by the outside world, the embedded watermark information will be destroyed. In order to solve the problem of distortion in robust watermarking technology and weak robustness of reversible watermarking technology, robust reversible watermarking is proposed. Through robust reversible watermarking technology, when the watermarked image is not attacked by the outside world, the decoder can extract the embedded watermark and reversibly restore the original image; if the watermarked image is attacked by the outside world, the decoder can extract the embedded watermark, but cannot restore the original image.

[0003] The two-stage watermarking method is an effective framework for the robust reversible watermarking method, which is divided into two stages. In the first stage, the watermark information is reasonably embedded into the host image using the corresponding robust watermarking method to obtain an intermediate image. This stage focuses on the watermark capacity and the distortion caused by watermark embedding. In the second stage, the information required to restore the original image from the intermediate image is calculated, and it is reasonably embedded into the intermediate image using the reversible watermarking method to obtain the final image containing the watermark information. In the decoder stage, if the image containing the watermark information is not attacked by the outside world, the corresponding watermark information is extracted and the original image is restored. If the watermark information is attacked by the outside world, only the watermark information can be extracted and the original image cannot be restored.

[0004] The existing two-stage watermarking method has the following problems:

[0005] 1. Robust watermark and reversible watermark are embedded in the same domain, and reversible watermark may distort the robust watermark.

[0006] 2. The watermark embedding domain is not robust enough, especially against geometric attacks.

[0007] 3. For the sake of image distortion, the watermark capacity of the proposed two-stage watermarking method is low.

[0008] 4. The proposed method has low imperceptibility. Summary of the invention

[0009] The present invention provides a high-capacity and low-distortion watermark method for copyright protection of digital artworks to overcome the above technical problems.

[0010] A high-capacity and low-distortion watermarking method for copyright protection of digital artworks, comprising:

[0011] S1, obtaining a digital artwork image I to be protected with a size of M×N, and embedding the robust watermark information into I to obtain a digital artwork image containing the robust watermark information;

[0012] S2, embedding the reversible watermark into the digital artwork image containing the robust watermark information to obtain the digital artwork image containing the robust watermark information and the reversible watermark;

[0013] S3, processing the digital artwork image containing the robust watermark information and the reversible watermark and sending it to the receiving end, and the receiving end verifies the integrity of the received digital artwork image containing the robust watermark information and the reversible watermark;

[0014] S4. If the receiving end passes the integrity verification, watermark extraction and image restoration are performed on the received digital artwork image containing robust watermark information and reversible watermark. Otherwise, watermark extraction is performed on the received digital artwork image containing robust watermark information and reversible watermark.

[0015] Preferably, the S1 includes:

[0016] S11, loading a digital artwork image I to be protected with a size of M×N;

[0017] S12, perform discrete inseparable shearlet transform on I to obtain a low-frequency subband I of size M×N low , determine the segmentation size to be a×a, and divide the low-frequency subband I according to the segmentation size low After segmentation, we get DNST coefficient blocks;

[0018] S13, according to the length L of the robust watermark sequence, using the first private key Key1 from DNST coefficient block selection DNST coefficient blocks are obtained by performing a T-order fast generalized polar coordinate complex exponential transformation on each selected DNST coefficient block. DNST-FGPCET amplitude blocks, each DNST-FGPCET amplitude block contains (2T+1) 2 DNST-FGPCET amplitudes, where the DNST-FGPCET amplitudes with order and repetition of 0 in each DNST-FGPCET amplitude block are not used to hide the robust watermark, and the remaining The DNST-FGPCET amplitude is used to hide the robust watermark sequence W,

[0019] Recorded as DF q,n,m represents the DNST-FGPCET amplitude with order n and repetition m in the qth DNST-FGPCET amplitude block;

[0020] S14, DF q,n,m After normalization, we get

[0021]

[0022] Among them, Hp is a hyperparameter, is the normalized DNST-FGPCET amplitude used to hide the watermark, DF q,0,0 Used to normalize the DNST-FGPCET amplitude, DF q,0,0 represents the DNST-FGPCET amplitude with order 0 and repetition 0 in the qth DNST-FGPCET amplitude block;

[0023] S15, using integer distortion compensation quantization index modulation method in After embedding the robust watermark in The details are as follows:

[0024]

[0025]

[0026]

[0027] where [·] is the rounding function, s=1,2,...,L, -T≤n≤T,-T≤m≤T,n≠0,m≠0,j={0,1},β s (j) is the jitter value and β s (1) = β s (0)+Δ / 2, Δ is the quantization step size, is the normalized DNST-FGPCET amplitude containing robust watermark information, is the normalized DNST-FGPCET amplitude rounding difference, is the robust watermark quantization error;

[0028] S16, yes Perform inverse normalization to obtain the DNST-FGPCET amplitude containing robust watermark information, as follows:

[0029]

[0030] S17, performing T-order FGPCET inverse calculation on the DNST-FGPCET amplitude containing the robust watermark information to obtain the DNST coefficient containing the robust watermark information, performing inverse DNST calculation on the DNST coefficient containing the robust watermark information and the DNST without the robust watermark information, and obtaining the digital artwork image I containing the robust watermark information rw .

[0031] Preferably, S2 includes:

[0032] S21, to I rw Perform DNST calculation to obtain the low-frequency subband containing robust watermark information Determine the amplitude of DNST-FGPCET with robust watermark information

[0033] S22, DNST-FGPCET amplitude with robust watermark information Perform the normalization operation to obtain According to formula (6), the DNST-FGPCET amplitude without robust watermark information is obtained

[0034]

[0035] S23, Perform inverse normalization to get

[0036] S24, yes Perform the inverse calculation of FGPCET to obtain the DNST coefficients of the de-robust watermark information, and perform the inverse calculation of DNST on the DNST coefficients of the de-robust watermark information to obtain

[0037] S25, through The rounding error D is calculated according to formula (7) r , construct compensation information set

[0038]

[0039] S26, using the second private key Key2 to obtain DNST coefficient block selection A DNST coefficient block is used to hide a reversible watermark;

[0040] S27, based on the prediction error expansion-histogram shift reversible watermarking method, hide C to the selected step S26 DNST coefficient blocks, we get the digital artwork image I containing robust watermark information and reversible watermark rrw .

[0041] Preferably, the S3 includes:

[0042] S31, to I rrw Perform a hash operation to obtain the first hash value H1, and then perform a hash operation on I according to the segmentation size a×a. rrw Perform the split, using the third private key Key3 from choose The DNST coefficient block without watermark information position is embedded with the first hash value H1 using the prediction error expansion-histogram shift reversible watermarking method. The DNST coefficient block without watermark information position is obtained to obtain the digital artwork image I containing the robust watermark, the reversible watermark and the first hash value. srrw ;

[0043] S32, I srrw After sending to the receiving end, Reversible watermarking method based on prediction error expansion-histogram shift Extract the embedded first hash value H1 from After removing the first hash value H1, we can recover calculate The second hash value H2, if H1 = H2, then it means No external attack, that is from Extract the watermark information and restore I, that is, the receiving end has passed the integrity verification; if H1≠H2, it means Under attack from outside, Only from The watermark information is extracted from the receiving end, which means that the receiving end fails the integrity verification.

[0044] Preferably, the S4 includes:

[0045] If the receiving end fails the integrity verification, watermark extraction is performed on the received digital artwork image containing the robust watermark information and the reversible watermark, wherein the watermark extraction comprises extracting the watermark from the received digital artwork image containing the robust watermark information and the reversible watermark according to steps S12 and S13. Select the one with robust watermark Extracting the watermark according to step S22;

[0046] On the contrary, watermark extraction and image restoration are performed on the received digital artwork image containing robust watermark information and reversible watermark, and the watermark extraction and image restoration on the received digital artwork image containing robust watermark information and reversible watermark includes:

[0047] S41, from Remove H1 to get Using prediction error expansion-histogram shift reversible watermarking method Extracting reversible watermark information set from Remove C ex Get a robust watermarked image

[0048] S42, from Select the one with robust watermark According to formula (1), normalization is performed to obtain According to formula (8) and (9), we can get all watermarks:

[0049]

[0050] S43, based on quantization error According to S21, S22, S23, S24, the de-robust watermark image is obtained Combined with rounding error D r Restore the digital artwork image I to be protected, as shown in formula (10):

[0051]

[0052] Among them, I cover Represents a restored digital artwork image.

[0053] The present invention provides a high-capacity, low-distortion watermarking method for copyright protection of digital artworks. The method combines discrete inseparable shearlet transform and local fast generalized polar coordinate complex exponential transform to perform robust watermarking and reversible watermarking embedding on digital artworks, thereby improving the imperceptibility, robustness and effective watermark capacity of the overall solution; the integrity of the digital artwork image embedded with the watermark is verified, and watermark extraction and image restoration are performed based on the integrity verification result, thereby effectively improving the copyright protection level of digital artworks. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0055] Figure 1 is a flow chart of the method of the present invention;

[0056] Figure 2 This is the processing flow of robust and reversible watermarking and integrity information verification embedding in an embodiment of the present invention;

[0057] Figure 3 It is a flow chart of integrity verification, watermark decoding and image restoration in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] Figure 1 It is a flow chart of the method of the present invention, such as Figure 1 As shown, the method of this embodiment includes:

[0060] S1, obtaining a digital artwork image I to be protected with a size of M×N, and embedding the robust watermark information into I to obtain a digital artwork image containing the robust watermark information;

[0061] S2, embedding the reversible watermark into the digital artwork image containing the robust watermark information to obtain the digital artwork image containing the robust watermark information and the reversible watermark;

[0062] S3, processing the digital artwork image containing the robust watermark information and the reversible watermark and sending it to the receiving end, and the receiving end verifies the integrity of the received digital artwork image containing the robust watermark information and the reversible watermark;

[0063] S4. If the receiving end passes the integrity verification, watermark extraction and image restoration are performed on the received digital artwork image containing robust watermark information and reversible watermark. Otherwise, watermark extraction is performed on the received digital artwork image containing robust watermark information and reversible watermark.

[0064] Based on the above scheme, the present invention provides a high-capacity, low-distortion watermarking method for copyright protection of digital artworks, which combines discrete inseparable shearlet transform and local fast generalized polar coordinate complex exponential transform to perform robust watermarking and reversible watermarking embedding on digital artworks, thereby improving the imperceptibility, robustness and effective watermark capacity of the overall scheme; performs integrity verification on the digital artwork image embedded with the watermark, and performs watermark extraction and image restoration based on the integrity verification result, thereby effectively improving the copyright protection level of digital artworks.

[0065] Specifically, this embodiment provides the details of the high-capacity and low-distortion watermarking method, and the processing flow of robust watermarking, reversible watermarking, and integrity information verification embedding is as follows: Figure 2 As shown in the figure, the flow chart of integrity verification, watermark decoding and image restoration is as follows Figure 3 As shown, including:

[0066] S1, obtaining a digital artwork image I to be protected with a size of M×N, and embedding the robust watermark information into I to obtain a digital artwork image containing the robust watermark information;

[0067] Specifically, the S1 includes:

[0068] S11, loading a digital artwork image I to be protected with a size of M×N;

[0069] S12, perform discrete non-separable shearlet transform (DNST) on I to a low-frequency subband I of size M×N low , determine the segmentation size to be a×a, and divide the low-frequency subband I according to the segmentation size low After segmentation, we get DNST coefficient blocks;

[0070] Among them, discrete inseparable shearlet transform is a mathematical tool used in signal processing and image analysis. It can be used for multi-resolution analysis and directional feature extraction. Since DNST is based on shearlet theory, it has good mathematical properties and can provide rich signal representation.

[0071] S13, according to the length L of the robust watermark sequence, using the customized first private key Key1 from DNST coefficient block selection DNST coefficient blocks are obtained by performing a T-order fast generalized polar coordinate complex exponential transformation on each selected DNST coefficient block. DNST-FGPCET amplitude blocks, each DNST-FGPCET amplitude block contains (2T+1) 2DNST-FGPCET amplitudes, where the DNST-FGPCET amplitudes with order and repetition of 0 in each DNST-FGPCET amplitude block are not used to hide the robust watermark, and the remaining The DNST-FGPCET amplitude is used to hide the robust watermark sequence W, denoted as DF q,n,m It represents the DNST-FGPCET amplitude with order n and repetition m in the qth DNST-FGPCET amplitude block, and DF refers to the amplitude set in these DNST-FGPCET amplitude blocks; wherein, Fast Generalized Polar Complex Exponential Transform (FGPCET) is a transformation method used for signal processing and image analysis, which has the characteristics of high computational efficiency and strong resistance to geometric attacks, and T refers to the order of FGPCET calculation.

[0072] S14, DF q,n,m After normalization, we get

[0073]

[0074] Among them, Hp is a hyperparameter, is the normalized DNST-FGPCET amplitude used to hide the watermark, using DF q,0,0 Used to normalize the DNST-FGPCET amplitude, DF q,0,0 Not embedded with watermark, DF q,0,0 represents the DNST-FGPCET amplitude with order 0 and repetition 0 in the qth target DNST-FGPCET amplitude block, and the target DNST-FGPCET amplitude block is the DNST-FGPCET amplitude block used to hide the watermark;

[0075] S15, Integer Distortion Compensated Quantization Index Modulation (IDC-QIM) After embedding the robust watermark in , as follows:

[0076]

[0077] Where [·] is the rounding function, s = 1 , 2,...,L, -T≤n≤T,-T≤m≤T,n≠0,m≠0,j={0,1},βs (j) is the jitter value and β s (1) = β s (0)+Δ / 2, Δ is the quantization step size, is the normalized DNST-FGPCET amplitude containing robust watermark information, is the normalized DNST-FGPCET amplitude rounding difference, is the robust watermark quantization error;

[0078] Among them, IDC-QIM is a quantization method used to embed information in a digital signal without significantly affecting the quality of the original signal. This method minimizes the impact of the embedded information on the signal by compensating for the distortion in the quantization process. This means that even if the information is embedded, the perceived quality of the signal remains high. IDC-QIM improves resistance to common attacks (such as compression, shearing, etc.), so that the embedded information can still be recovered after a certain degree of signal processing. Through the quantization method, IDC-QIM is able to embed different types of information in different parts of the signal, enhancing the flexibility and efficiency of information embedding.

[0079] S16, yes Perform inverse normalization to obtain the DNST-FGPCET amplitude containing robust watermark information, as follows:

[0080]

[0081] S17, performing T-order FGPCET inverse calculation on the DNST-FGPCET amplitude containing the robust watermark information to obtain the DNST coefficient containing the robust watermark information, performing inverse DNST calculation on the DNST coefficient containing the robust watermark information and the DNST without the robust watermark information, and obtaining the digital artwork image I containing the robust watermark information rw .

[0082] S2, embedding the reversible watermark into the digital artwork image containing the robust watermark information to obtain the digital artwork image containing the robust watermark information and the reversible watermark;

[0083] Specifically, the S2 includes:

[0084] S21, to I rw Perform DNST calculation to obtain the low-frequency subband containing robust watermark information Determine the amplitude of DNST-FGPCET with robust watermark information

[0085] S22, DNST-FGPCET amplitude with robust watermark information Perform the normalization operation to obtain According to formula (6), the DNST-FGPCET amplitude without robust watermark information is obtained

[0086]

[0087] S23, Perform inverse normalization to get

[0088] S24, yes Perform the inverse calculation of FGPCET to obtain the DNST coefficients of the de-robust watermark information, and perform the inverse calculation of DNST on the DNST coefficients of the de-robust watermark information to obtain

[0089] S25, through The rounding error D is calculated according to formula (7) r , construct compensation information set

[0090]

[0091] S26, using the second private key Key2 to obtain DNST coefficient block selection A DNST coefficient block is used to hide a reversible watermark;

[0092] S27, based on the prediction error expansion-histogram shifting (PEE-HS) reversible watermarking method, hide C to the selected in step S26 DNST coefficient blocks, we get the digital artwork image I containing robust watermark information and reversible watermark rrw .

[0093] S3, processing the digital artwork image containing the robust watermark information and the reversible watermark and sending it to the receiving end, and the receiving end verifies the integrity of the received digital artwork image containing the robust watermark information and the reversible watermark;

[0094] Specifically, the S3 includes:

[0095] S31, to I rrw Perform a hash operation to obtain the first hash value H1, and then perform a hash operation on I according to the segmentation size a×a. rrw Perform the split, using the third private key Key3 from choose The DNST coefficient block without watermark information position is embedded with the first hash value H1 using the prediction error expansion-histogram shift reversible watermarking method. The DNST coefficient block without watermark information position is obtained to obtain the digital artwork image I containing the robust watermark, the reversible watermark and the first hash value. srrw ;

[0096] S32. Since the image may be tampered or the data may be lost when it is transmitted over the network, srrw After sending to the receiving end, Reversible watermarking method based on prediction error expansion-histogram shift Extract the embedded first hash value H1 from After removing the first hash value H1, we can recover calculate The second hash value H2, if H1 = H2, then it means No external attack, that is from Extract the watermark information and restore I, that is, the receiving end has passed the integrity verification; if H1≠H2, it means Under attack from outside, Only from The watermark information is extracted from the receiving end, which means that the receiving end fails the integrity verification.

[0097] S4. If the receiving end passes the integrity verification, watermark extraction and image restoration are performed on the received digital artwork image containing robust watermark information and reversible watermark. Otherwise, watermark extraction is performed on the received digital artwork image containing robust watermark information and reversible watermark.

[0098] Specifically, the S4 includes:

[0099] If the receiving end fails the integrity verification, watermark extraction is performed on the received digital artwork image containing the robust watermark information and the reversible watermark, wherein the watermark extraction comprises extracting the watermark from the received digital artwork image containing the robust watermark information and the reversible watermark according to steps S12 and S13. Select the one with robust watermark Extracting the watermark according to step S22;

[0100] On the contrary, watermark extraction and image restoration are performed on the received digital artwork image containing robust watermark information and reversible watermark, and the watermark extraction and image restoration on the received digital artwork image containing robust watermark information and reversible watermark includes:

[0101] S41, from Remove H1 to get , using PEE-HS from Extracting reversible watermark information set from Remove Cex Get a robust watermarked image

[0102] S42, from Select the one with robust watermark According to formula (1), normalization is performed to obtain based on And robust watermark decoding to get the whole watermark

[0103]

[0104] S43, based on quantization error According to S21, S22, S23, S24, the de-robust watermark image is obtained Combined with rounding error D r Restore the digital artwork image I to be protected, as shown in formula (10):

[0105]

[0106] Among them, I cover Represents a restored digital artwork image.

[0107] Overall beneficial effects:

[0108] The present invention provides a high-capacity, low-distortion watermarking method for copyright protection of digital artworks, which combines discrete inseparable shearlet transform and local fast generalized polar coordinate complex exponential transform to perform robust watermarking and reversible watermarking embedding, thereby improving the imperceptibility, robustness and effective watermark capacity of the overall solution; and improves the copyright protection level of digital artworks based on image integrity verification and watermark extraction and restoration.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A high-capacity and low-distortion watermarking method for copyright protection of digital artworks. It is characterized in that include: S1, obtaining a digital artwork image I to be protected with a size of M×N, and embedding the robust watermark information into I to obtain a digital artwork image containing the robust watermark information; S2, embedding the reversible watermark into the digital artwork image containing the robust watermark information to obtain the digital artwork image containing the robust watermark information and the reversible watermark; S3, processing the digital artwork image containing the robust watermark information and the reversible watermark and sending it to the receiving end, and the receiving end verifies the integrity of the received digital artwork image containing the robust watermark information and the reversible watermark; S4. If the receiving end passes the integrity verification, watermark extraction and image restoration are performed on the received digital artwork image containing robust watermark information and reversible watermark. Otherwise, watermark extraction is performed on the received digital artwork image containing robust watermark information and reversible watermark.

2. A high-capacity and low-distortion watermarking method for copyright protection of digital artworks according to claim 1, It is characterized in that The S1 includes: S11, loading a digital artwork image I to be protected with a size of MΔN; S12, perform discrete inseparable shearlet transform on I to obtain a low-frequency subband I of size M×N low , determine the segmentation size to be a×a, and divide the low-frequency subband I according to the segmentation size low After segmentation, we get DNST coefficient blocks; S13, according to the length L of the robust watermark sequence, using the first private key Key 1 from DNST coefficient block selection DNST coefficient blocks are obtained by performing a T-order fast generalized polar coordinate complex exponential transformation on each selected DNST coefficient block. DNST-FGPCET amplitude blocks, each DNST-FGPCET amplitude block contains (2T+1) 2 DNST-FGPCET amplitudes, where the DNST-FGPCET amplitudes with order and repetition of 0 in each DNST-FGPCET amplitude block are not used to hide the robust watermark, and the remaining The DNST-FGPCET amplitude is used to hide the robust watermark sequence W, Recorded as DF q,n,m represents the DNST-FGPCET amplitude with order n and repetition m in the qth DNST-FGPCET amplitude block; S14, DF q,n,m After normalization, we get Among them, Hp is a hyperparameter, is the normalized DNST-FGPCET amplitude used to hide the watermark, DF q,0,0 Used to normalize the DNST-FGPCET amplitude, DF q,0,0 represents the DNST-FGPCET amplitude with order 0 and repetition 0 in the qth DNST-FGPCET amplitude block; S15, using integer distortion compensation quantization index modulation method in After embedding the robust watermark in The details are as follows: where [·] is the rounding function, s=1,2,...,L, -T≤n≤T,-T≤m≤T,n≠0,m≠0,j={0,1},β s (j) is the jitter value and β s (1) = β s (0)+Δ / 2, Δ is the quantization step size, is the normalized DNST-FGPCET amplitude containing robust watermark information, is the normalized DNST-FGPCET amplitude rounding difference, is the robust watermark quantization error; S16, yes Perform inverse normalization to obtain the DNST-FGPCET amplitude containing robust watermark information, as follows: S17, performing T-order FGPCET inverse calculation on the DNST-FGPCET amplitude containing the robust watermark information to obtain the DNST coefficient containing the robust watermark information, performing inverse DNST calculation on the DNST coefficient containing the robust watermark information and the DNST without the robust watermark information, and obtaining the digital artwork image I containing the robust watermark information rw .

3. A high-capacity and low-distortion watermarking method for copyright protection of digital artworks according to claim 2, It is characterized in that The S2 includes: S21, to I rw Perform DNST calculation to obtain the low-frequency subband containing robust watermark information To determine the DNST-FGPCET amplitude containing robust watermark information S22, DNST-FGPCET amplitude with robust watermark information Perform the normalization operation to obtain According to formula (6), the DNST-FGPCET amplitude without robust watermark information is obtained S23, Perform inverse normalization to get S24, yes Perform the inverse calculation of FGPCET to obtain the DNST coefficients of the de-robust watermark information, and perform the inverse calculation of DNST on the DNST coefficients of the de-robust watermark information to obtain S25, through The rounding error D is calculated according to formula (7) r , construct compensation information set S26. Using the second private key Key 2 From other DNST coefficient block selection A DNST coefficient block is used to hide a reversible watermark; S27, based on the prediction error expansion-histogram shift reversible watermarking method, hide C to the selected step S26 DNST coefficient blocks, we get the digital artwork image I containing robust watermark information and reversible watermark rrw .

4. A high-capacity and low-distortion watermarking method for copyright protection of digital artworks according to claim 3, It is characterized in that The S3 includes: S31, to I rrw Perform a hash operation to obtain the first hash value H 1 , according to the segmentation size a×a for I rrw Perform the split, using the third private key Key 3 from choose DNST coefficient blocks without watermark information are used to convert the first hash value H 1 Embed The DNST coefficient block without watermark information position is obtained to obtain the digital artwork image I containing the robust watermark, the reversible watermark and the first hash value. srrw ; S32, I srrw After sending to the receiving end, Reversible watermarking method based on prediction error expansion-histogram shift Extract the embedded first hash value H 1 ,from Remove the first hash value H 1 After recovery calculate The second hash value H 2 , if H 1 =H 2 , then it means No external attack, that is from Extract the watermark information and restore I, that is, the receiving end has passed the integrity verification; if H 1 ≠H 2 , then it means Under attack from outside, Only from The watermark information is extracted from the receiving end, which means that the receiving end fails the integrity verification.

5. A high-capacity and low-distortion watermarking method for copyright protection of digital artworks according to claim 4, It is characterized in that The S4 includes: If the receiving end fails the integrity verification, watermark extraction is performed on the received digital artwork image containing the robust watermark information and the reversible watermark, wherein the watermark extraction comprises extracting the watermark from the received digital artwork image containing the robust watermark information and the reversible watermark according to steps S12 and S13. Select the one with robust watermark Extracting the watermark according to step S22; On the contrary, watermark extraction and image restoration are performed on the received digital artwork image containing robust watermark information and reversible watermark, and the watermark extraction and image restoration on the received digital artwork image containing robust watermark information and reversible watermark includes: S41, from Remove H 1 get Using prediction error expansion-histogram shift reversible watermarking method Extracting reversible watermark information set from Remove C ex Get a robust watermarked image S42, from Select the one with robust watermark According to formula (1), we can get According to formula (8) and (9), we can get all watermarks: S43, based on quantization error According to S21, S22, S23, S24, the de-robust watermark image is obtained Combined with rounding error D r Restore the digital artwork image I to be protected, as shown in formula (10): Among them, I cover Represents a restored digital artwork image.

Images