A digital watermark camouflage method based on reverse information entropy tracing

Abstract

The present invention discloses a digital watermark camouflage method based on reverse information entropy tracking, comprising the following steps: dynamically adjusting the embedding strength and concealment of a digital watermark by monitoring the fluctuation of the local entropy value of a signal at different processing stages; monitoring the complexity of a signal by local entropy changes during signal processing, and adjusting the camouflage degree of a digital watermark according to the amount of change; adjusting key parameters in the extraction and recovery process of a digital watermark in real time by an information entropy feedback mechanism, optimizing the recovery accuracy of a digital watermark; optimizing the embedding process of a digital watermark by using a reverse information entropy minimization method in a high noise environment, and improving the anti-interference capability of a digital watermark under noise interference conditions. The present invention ensures effective recovery and stable transmission of digital watermark information under various noise and signal change conditions by dynamically optimizing the embedding and recovery strategies of digital watermarks.

Description

Technical Field

[0001] The present invention relates to the technical field of digital watermarking, and in particular to a digital watermark camouflage method based on reverse information entropy tracing. Background Art

[0002] With the widespread application of digital watermarking technology, watermark embedding and extraction methods play an important role in copyright protection, information authentication, data privacy and other fields. With the continuous development of transmission channel environment and signal processing technology, watermarking technology is facing increasingly complex challenges, especially in environments with strong noise interference. Traditional watermarking methods often find it difficult to balance concealment and anti-interference, which makes the watermark information vulnerable to damage, affecting the effectiveness and recovery accuracy of the watermark.

[0003] In the prior art, digital watermarking technology usually adopts fixed embedding strategies and recovery algorithms. These methods mostly rely on static analysis of local signal features, ignoring the dynamic changes of signals at different processing stages. Therefore, they are difficult to flexibly adjust the watermark embedding and extraction process when facing a high noise environment, and cannot fully optimize the robustness of the watermark. Specifically, the prior art has the following shortcomings:

[0004] 1. Lack of dynamic adaptability: Traditional watermarking technology fails to fully consider the dynamic changes of signals and lacks a real-time entropy value change tracking mechanism, resulting in the inability of watermark embedding and recovery processes to adapt to the complex changes in different signal environments.

[0005] 2. Inefficient noise processing capability: Existing methods have poor adaptability to noise interference and cannot effectively eliminate the impact of noise on watermark recovery, resulting in the loss or inability to recover watermark information in a noisy environment.

[0006] 3. Static embedding and extraction strategies: Existing watermark embedding methods usually adopt fixed strategies, ignoring the combined effects of signal characteristics and noise influences, and are unable to optimize the watermark recovery accuracy and concealment.

[0007] 4. Lack of signal feedback mechanism: The existing technology fails to adopt a dynamic feedback mechanism based on entropy value changes in the watermark recovery process, and cannot adjust the watermark recovery process according to real-time signal changes, which affects the improvement of recovery accuracy.

[0008] Therefore, how to provide a digital watermark camouflage method based on reverse information entropy tracing is a problem that those skilled in the art need to solve urgently. Summary of the invention

[0009] One purpose of the present invention is to propose a digital watermark camouflage method based on reverse information entropy tracking. The present invention combines information entropy adaptive adjustment and dynamic feedback mechanism, and optimizes the digital watermark embedding and recovery process by utilizing the comprehensive effect of local entropy change and signal characteristics. Through dynamic monitoring and adjustment of entropy value changes, the present invention can effectively improve the watermark recovery accuracy and anti-interference ability in complex signal environments, and has the advantages of high efficiency, flexibility and strong robustness.

[0010] A digital watermark camouflage method based on reverse information entropy tracing according to an embodiment of the present invention comprises the following steps:

[0011] S1. Monitor the entropy changes of the signal to be embedded, obtain the local entropy value of the signal through the reverse information entropy tracking mechanism, and record the entropy fluctuations of the signal at different processing stages;

[0012] S2, according to the monitored entropy value change of the signal to be embedded and the local statistical characteristics of the signal, the entropy value gradient is calculated to determine the specific location, timing and strength of the digital watermark embedding;

[0013] S3, optimize the digital watermark embedding process through the entropy gradient, dynamically adjust the distribution and strength of the digital watermark, so that the digital watermark and signal characteristics can be organically integrated;

[0014] S4, according to the adjusted digital watermark distribution, monitor the fluctuation of the signal entropy value in real time, use the entropy feedback mechanism to adjust the camouflage degree of the digital watermark, and control the detectability of the digital watermark at different signal processing stages;

[0015] S5. Based on the entropy value change of the entropy value feedback mechanism, the information entropy adaptive adjustment method is used to dynamically correct the digital watermark extraction and recovery process, so as to achieve effective recovery of digital watermark information under signal changes and noise interference;

[0016] S6. In a high noise environment, combined with the adjusted digital watermark information, the inverse information entropy minimization method is used to optimize the digital watermark embedding process and the anti-interference ability during signal transmission and processing.

[0017] Optionally, the S1 specifically includes:

[0018] S11. Apply the reverse information entropy tracking mechanism to calculate the local entropy value of the embedded signal. The local entropy value represents the statistical characteristics of the signal at a specific processing stage and measures the complexity and uncertainty of the signal at this stage.

[0019] S12, defining a time window for entropy value calculation, wherein the time window corresponds to each processing stage of the signal to be embedded, and in each stage, obtaining a local entropy value change of the stage through real-time sampling and calculation;

[0020] S13. According to the local statistical characteristics of the signal to be embedded, select a suitable entropy value calculation method, wherein the entropy value calculation method is based on the Shannon entropy formula:

[0021] ;

[0022] in, is the local entropy value of signal X in time window t, is the probability distribution of the i-th data point of signal X in time window t;

[0023] S14, based on the calculated local entropy value , monitoring and recording the entropy fluctuations of the signal at different processing stages, wherein the fluctuations measure the magnitude of the change of the signal at each stage;

[0024] S15. By analyzing the recorded entropy fluctuations, the signal segment with the smallest entropy fluctuation is identified as the target area for digital watermark embedding, and according to the entropy fluctuations of the signal at each stage, the entropy threshold range is set to perform quantitative analysis.

[0025] Optionally, the S3 specifically includes:

[0026] S31, based on the calculated entropy value gradient and in combination with the local statistical characteristics of the signal to be embedded, determining the digital watermark embedding strength of different signal regions, and adjusting the digital watermark embedding strength according to the change of the local entropy gradient;

[0027] S32. Calculate the corresponding digital watermark embedding strength for different regions of the signal through the obtained local entropy gradient:

[0028] ;

[0029] in, represents the digital watermark embedding strength at position x, is the digital watermark strength adjustment coefficient, is the absolute value of the local entropy gradient at position x, represents the local statistical characteristics of the signal at position x, Indicates the signal characteristics at that location;

[0030] S33, based on the calculated digital watermark embedding strength , select the priority embedding area of ​​the signal, the priority embedding area is the area with higher entropy gradient than other areas, and determine the digital watermark embedding position based on the complexity of the area with higher entropy gradient than other areas and the change of entropy gradient;

[0031] S34, based on the selected priority embedding area and the calculated digital watermark embedding strength , the embedding position and strength of the digital watermark are adjusted, and the digital watermark information and the local characteristics of the signal are weightedly distributed to achieve the fusion of the digital watermark and the signal;

[0032] S35. Monitor the entropy gradient change during the signal processing process, dynamically optimize the digital watermark embedding process according to the current entropy gradient of the signal, and adjust the digital watermark embedding strategy so that the digital watermark can achieve the desired fusion effect at each processing stage of the signal.

[0033] Optionally, the S4 specifically includes:

[0034] S41, monitoring and recording the fluctuation of the local entropy value of the signal at different processing stages, extracting the change rate of the local entropy value of each signal segment, and determining the complexity change of the signal based on the change in the entropy value;

[0035] S42, defining and calculating the entropy value fluctuation threshold of each signal processing stage, and triggering the dynamic adjustment mechanism of the digital watermark camouflage degree when the change rate of the local entropy value exceeds the threshold;

[0036] S43. According to the monitored local entropy value change rate, a dynamic feedback relationship is established to link the entropy value change with the adjustment function of the digital watermark camouflage degree, which is used to quantify the camouflage degree adjustment:

[0037] ;

[0038] in, It represents the change of the local entropy value of the signal at position x between time periods t and t+1. is the adjustment coefficient;

[0039] S44, based on the feedback entropy value change, by weighted average signal characteristics and entropy value change, calculate the adjustment amount of the digital watermark camouflage degree in each area:

[0040] ;

[0041] in, represents the camouflage degree of the digital watermark at position x, is the adjustment weight of the entropy value change on the camouflage degree, is the adjustment weight of signal characteristics on camouflage degree, Represents the statistical characteristics of the signal at position x;

[0042] S45. Based on the calculated digital watermark camouflage adjustment value, the digital watermark embedding strategy is dynamically optimized, and the embedding strength and concealment of the digital watermark at different signal processing stages are controlled in real time, so that the detectability of the digital watermark during signal transmission is maintained within the set threshold.

[0043] Optionally, the S5 specifically includes:

[0044] S51, monitoring the change of the local entropy value of the signal in the digital watermark extraction stage, generating an entropy value feedback signal and performing real-time quantization of the signal characteristics of the digital watermark extraction based on the signal;

[0045] S52, combining the entropy feedback signal, adjusting the key parameters in the digital watermark extraction process through the information entropy adaptive adjustment method, optimizing the extraction accuracy by using the dynamic characteristics of the entropy value change, and adapting to the influence of signal changes and noise interference;

[0046] S53, based on the entropy feedback mechanism, a dynamic adjustment strategy based on local entropy changes is adopted to automatically correct the errors in the digital watermark recovery process, so that the accuracy and robustness of the digital watermark recovery process can be improved under the conditions of signal noise interference and signal characteristic changes;

[0047] S54, constructing a mapping relationship between local entropy change and digital watermark recovery accuracy, and constructing a digital watermark recovery optimization function through the mapping relationship, which adjusts the recovery strategy according to the entropy value change and signal characteristics of the position x:

[0048] ;

[0049] in, represents the optimal value of digital watermark recovery at position x, is the adaptive adjustment coefficient, is the local entropy change at position x, Represents the local statistical characteristics of the location, is the signal characteristic adjustment coefficient;

[0050] S55. Using the recovery adjustment value generated by the digital watermark recovery optimization function, dynamically correct the parameters in the digital watermark extraction and recovery process, so that during the signal transmission process, the digital watermark can be effectively recovered under different noise interference conditions, while ensuring that the concealment and detectability of the digital watermark information remain within a preset threshold.

[0051] Optionally, the S52 specifically includes:

[0052] S521, monitor the change of the local entropy value of the signal in the digital watermark extraction stage, calculate and generate the local entropy change rate of each signal segment , which is used to quantify the dynamic characteristics of the signal in the extraction stage. The entropy change rate is calculated by the difference between the local entropy value of the current signal segment and the local entropy value of the previous signal segment, and then normalized:

[0053] ;

[0054] in, represents the local entropy value at position x at time t, Represents the local entropy value at position x at time t-1;

[0055] S522, based on the local entropy change rate , extract the noise interference characteristics of the signal, and use the noise influence factor To quantify the noise interference of the signal, the noise impact factor is determined by the entropy change rate and the variance of the signal noise:

[0056] ;

[0057] in, is the noise impact factor at position x, and is the adjustment coefficient, is the variance of the signal noise at position x;

[0058] S523, combined with the local entropy change rate and noise impact factor ,Through the information entropy adaptive adjustment method, the key parameters in the digital watermark extraction process are adjusted to optimize the extraction accuracy. The adjustment strategy optimizes the noise suppression and accuracy control in the extraction process according to the dynamic entropy characteristics:

[0059] ;

[0060] in, represents the extraction process adjustment value at position x, and is the adjustment coefficient;

[0061] S524, according to the adjustment function Dynamically adjust key parameters in the digital watermark extraction process, including noise suppression strength, extraction accuracy and digital watermark extraction strategy. The specific extraction strategy includes selecting adaptive filtering algorithm, optimizing digital watermark extraction algorithm, and dynamically adjusting the extraction process in combination with local signal characteristics.

[0062] S525. Utilize the adjusted key parameters of the extraction process to optimize the digital watermark extraction process in real time, optimize its adaptability to different noise interferences and signal changes, and enable the digital watermark to be restored stably and effectively during signal transmission.

[0063] Optionally, the S54 specifically includes:

[0064] S541, based on the local entropy change at position x , construct the mapping relationship between local entropy change and digital watermark recovery accuracy, which quantifies the impact of local entropy change on digital watermark recovery accuracy;

[0065] S542, according to the local entropy change at position x And the signal statistical characteristics of the location , construct a digital watermark recovery optimization function, which is used to dynamically adjust the digital watermark recovery strategy and adaptively adjust the parameters of the digital watermark recovery process according to the comprehensive effect of local entropy changes and signal characteristics:

[0066] ;

[0067] in, represents the optimal value of digital watermark recovery at position x, is a constant offset;

[0068] S543, digital watermark recovery optimization function ,determine the key parameter adjustment strategy in the digital watermark recovery process, and dynamically adjust the parameters in the digital watermark recovery process through real-time feedback of local entropy changes and signal characteristics to achieve efficient recovery optimization;

[0069] S544, restore the optimization function based on the digital watermark The output of the adaptive correction is used to modify the control parameters in the recovery process. This correction mechanism is based on the dynamic response of local entropy changes and signal characteristics, so that the digital watermark recovery process can maintain the optimal recovery accuracy and stability under different signal conditions.

[0070] S545. By optimizing the restoration adjustment parameters, various key operations in the digital watermark restoration process are dynamically corrected, so that the digital watermark information can be restored stably and effectively under the conditions of signal noise interference and signal changes.

[0071] Optionally, the S6 specifically includes:

[0072] S61. In a high noise environment, monitor the noise characteristics during signal transmission, and combine the adjusted digital watermark information to optimize the digital watermark embedding process using the inverse information entropy minimization method. By quantifying the impact of signal noise, the noise characteristics are evaluated using the change in local entropy value, and the parameters of the digital watermark embedding process are adjusted to optimize the embedding strategy of the digital watermark information.

[0073] S62, based on the local entropy change at position x And the signal statistical characteristics of the location , construct the optimization objective function in the digital watermark embedding process, and quantify the impact of noise on the accuracy of digital watermark recovery:

[0074] ;

[0075] in, represents the optimization objective function, is the local entropy change at position x, is the noise impact factor at position x, is the adjustment coefficient;

[0076] S63, according to the optimization objective function , dynamically adjust the key parameters in the digital watermark embedding process, and feedback the local entropy changes in real time and noise impact factor , adjust the digital watermark embedding strength, embedding position and digital watermark format to optimize the digital watermark's anti-interference ability;

[0077] S64, based on the optimization objective function The output of is used to adjust the control parameters in the digital watermark embedding process. The adjustment is based on the dynamic response of local entropy changes and signal noise to optimize the digital watermark embedding accuracy and strength, so that the digital watermark maintains the optimal recovery accuracy and stability under different signal conditions.

[0078] S65. In the process of signal transmission, the key parameters of the digital watermark embedding process are dynamically adjusted by combining the inverse information entropy minimization method with the optimized digital watermark embedding strategy. The embedding parameters are adjusted by the output of the signal generator, so that the digital watermark has anti-interference ability during signal transmission and can be effectively restored under noisy conditions.

[0079] The beneficial effects of the present invention are:

[0080] The present invention combines reverse information entropy tracking and adaptive adjustment methods, utilizes the dynamic feedback mechanism of local entropy changes and signal characteristics, optimizes the embedding and extraction process of digital watermarks, and enhances the watermark recovery accuracy and anti-interference ability in complex signal environments. This method can effectively deal with signal noise interference and signal changes, and ensure that digital watermark information is stably and effectively restored during signal transmission.

[0081] The present invention dynamically optimizes the key parameters in the watermark recovery process by real-time monitoring and adjusting local entropy changes and signal characteristics, thereby improving the anti-interference ability of the watermark in a high-noise environment. Through the reverse information entropy minimization optimization algorithm, the present invention can flexibly adjust the watermark embedding strategy under different signal conditions, significantly improving the concealment and recoverability of the watermark.

[0082] The present invention proposes an adaptive watermark recovery optimization function by constructing a mapping relationship between local entropy changes and watermark recovery accuracy, so that the watermark recovery process can maintain the best recovery accuracy under different noise interference and signal change environments. This method avoids the static processing limitations of traditional watermarking technology and has high dynamic adaptability and robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0084] Figure 1 A flowchart of a digital watermark camouflage method based on reverse information entropy tracing proposed by the present invention;

[0085] Figure 2 It is a schematic diagram of constructing a watermark recovery optimization function based on the mapping relationship between local entropy change and watermark recovery accuracy in the present invention. DETAILED DESCRIPTION

[0086] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0087] refer to Figure 1-2 , a digital watermark camouflage method based on reverse information entropy tracing, comprising the following steps:

[0088] S1. Monitor the entropy changes of the signal to be embedded, obtain the local entropy value of the signal through the reverse information entropy tracking mechanism, and record the entropy fluctuations of the signal at different processing stages;

[0089] S2, according to the monitored entropy value change of the signal to be embedded and the local statistical characteristics of the signal, the entropy value gradient is calculated to determine the specific location, timing and strength of the digital watermark embedding;

[0090] S3, optimize the digital watermark embedding process through the entropy gradient, dynamically adjust the distribution and strength of the watermark, so that the digital watermark and signal characteristics can be organically integrated;

[0091] S4, according to the adjusted digital watermark distribution, monitor the fluctuation of the signal entropy value in real time, use the entropy feedback mechanism to adjust the camouflage degree of the digital watermark, and control the detectability of the digital watermark at different signal processing stages;

[0092] S5. Based on the entropy value change of the entropy value feedback mechanism, the information entropy adaptive adjustment method is used to dynamically correct the digital watermark extraction and recovery process, so as to achieve effective recovery of digital watermark information under signal changes and noise interference;

[0093] S6. In a high noise environment, combined with the adjusted digital watermark information, the inverse information entropy minimization method is used to optimize the digital watermark embedding process and the anti-interference ability during signal transmission and processing.

[0094] In this implementation, S1 specifically includes:

[0095] S11. Apply the reverse information entropy tracking mechanism to calculate the local entropy value of the embedded signal. The local entropy value represents the statistical characteristics of the signal at a specific processing stage and measures the complexity and uncertainty of the signal at this stage.

[0096] S12, defining a time window for entropy value calculation, wherein the time window corresponds to each processing stage of the signal to be embedded, and in each stage, obtaining a local entropy value change of the stage through real-time sampling and calculation;

[0097] S13. According to the local statistical characteristics of the signal to be embedded, select a suitable entropy value calculation method, wherein the entropy value calculation method is based on the Shannon entropy formula:

[0098] ;

[0099] in, is the local entropy value of signal X in time window t, is the probability distribution of the i-th data point of signal X in time window t;

[0100] S14, based on the calculated local entropy value , monitoring and recording the entropy fluctuations of the signal at different processing stages, wherein the fluctuations measure the magnitude of the change of the signal at each stage;

[0101] S15. By analyzing the recorded entropy fluctuations, the signal segment with the smallest entropy fluctuation is identified as the target area for digital watermark embedding, and according to the entropy fluctuations of the signal at each stage, the entropy threshold range is set to perform quantitative analysis.

[0102] In this implementation, S3 specifically includes:

[0103] S31, based on the calculated entropy value gradient and in combination with the local statistical characteristics of the signal to be embedded, determining the digital watermark embedding strength of different signal regions, and adjusting the digital watermark embedding strength according to the change of the local entropy gradient;

[0104] S32. Calculate the corresponding digital watermark embedding strength for different regions of the signal through the obtained local entropy gradient:

[0105] ;

[0106] in, represents the digital watermark embedding strength at position x, is the digital watermark strength adjustment coefficient, is the absolute value of the local entropy gradient at position x, represents the local statistical characteristics of the signal at position x, Indicates the signal characteristics at that location;

[0107] S33, based on the calculated digital watermark embedding strength , select the priority embedding area of ​​the signal, the priority embedding area is the area with higher entropy gradient than other areas, and determine the digital watermark embedding position based on the complexity of the area with higher entropy gradient than other areas and the change of entropy gradient;

[0108] S34, based on the selected priority embedding area and the calculated digital watermark embedding strength , the embedding position and strength of the digital watermark are adjusted, and the digital watermark information and the local characteristics of the signal are weightedly distributed to achieve the fusion of the digital watermark and the signal;

[0109] S35. Monitor the entropy gradient change during the signal processing process, dynamically optimize the digital watermark embedding process according to the current entropy gradient of the signal, and adjust the digital watermark embedding strategy so that the digital watermark can achieve the desired fusion effect at each processing stage of the signal.

[0110] In this implementation, S4 specifically includes:

[0111] S41, monitoring and recording the fluctuation of the local entropy value of the signal at different processing stages, extracting the change rate of the local entropy value of each signal segment, and determining the complexity change of the signal based on the change in the entropy value;

[0112] S42, defining and calculating the entropy value fluctuation threshold of each signal processing stage, and triggering the dynamic adjustment mechanism of the digital watermark camouflage degree when the change rate of the local entropy value exceeds the threshold;

[0113] S43. According to the monitored local entropy value change rate, a dynamic feedback relationship is established to link the entropy value change with the adjustment function of the digital watermark camouflage degree, which is used to quantify the camouflage degree adjustment:

[0114] ;

[0115] in, It represents the change of the local entropy value of the signal at position x between time periods t and t+1. is the adjustment coefficient;

[0116] S44, based on the feedback entropy value change, by weighted average signal characteristics and entropy value change, calculate the adjustment amount of the digital watermark camouflage degree in each area:

[0117] ;

[0118] in, represents the camouflage degree of the digital watermark at position x, is the adjustment weight of the entropy value change on the camouflage degree, is the adjustment weight of signal characteristics on camouflage degree, Represents the statistical characteristics of the signal at position x;

[0119] S45. Based on the calculated digital watermark camouflage adjustment value, the digital watermark embedding strategy is dynamically optimized, and the embedding strength and concealment of the digital watermark at different signal processing stages are controlled in real time, so that the detectability of the digital watermark during signal transmission is maintained within the set threshold.

[0120] In this implementation manner, S5 specifically includes:

[0121] S51, monitoring the change of the local entropy value of the signal in the digital watermark extraction stage, generating an entropy value feedback signal and performing real-time quantization of the signal characteristics of the digital watermark extraction based on the signal;

[0122] S52, combining the entropy feedback signal, adjusting the key parameters in the digital watermark extraction process through the information entropy adaptive adjustment method, optimizing the extraction accuracy by using the dynamic characteristics of the entropy value change, and adapting to the influence of signal changes and noise interference;

[0123] S53, based on the entropy feedback mechanism, a dynamic adjustment strategy based on local entropy changes is adopted to automatically correct the errors in the digital watermark recovery process, so that the accuracy and robustness of the digital watermark recovery process can be improved under the conditions of signal noise interference and signal characteristic changes;

[0124] S54, constructing a mapping relationship between local entropy change and digital watermark recovery accuracy, and constructing a digital watermark recovery optimization function through the mapping relationship, which adjusts the recovery strategy according to the entropy value change and signal characteristics of the position x:

[0125] ;

[0126] in, represents the optimal value of digital watermark recovery at position x, is the adaptive adjustment coefficient, is the local entropy change at position x, Represents the local statistical characteristics of the location, is the signal characteristic adjustment coefficient;

[0127] S55. Using the recovery adjustment value generated by the digital watermark recovery optimization function, dynamically correct the parameters in the digital watermark extraction and recovery process, so that during the signal transmission process, the digital watermark can be effectively recovered under different noise interference conditions, while ensuring that the concealment and detectability of the digital watermark information remain within a preset threshold.

[0128] In this implementation manner, the S52 specifically includes:

[0129] S521, monitor the change of the local entropy value of the signal in the digital watermark extraction stage, calculate and generate the local entropy change rate of each signal segment , which is used to quantify the dynamic characteristics of the signal in the extraction stage. The entropy change rate is calculated by the difference between the local entropy value of the current signal segment and the local entropy value of the previous signal segment, and then normalized:

[0130] ;

[0131] in, represents the local entropy value at position x at time t, Represents the local entropy value at position x at time t-1;

[0132] S522, based on the local entropy change rate , extract the noise interference characteristics of the signal, and use the noise influence factor To quantify the noise interference of the signal, the noise impact factor is determined by the entropy change rate and the variance of the signal noise:

[0133] ;

[0134] in, is the noise impact factor at position x, and is the adjustment coefficient, is the variance of the signal noise at position x;

[0135] S523, combined with the local entropy change rate and noise impact factor ,Through the information entropy adaptive adjustment method, the key parameters in the digital watermark extraction process are adjusted to optimize the extraction accuracy. The adjustment strategy optimizes the noise suppression and accuracy control in the extraction process according to the dynamic entropy characteristics:

[0136] ;

[0137] in, represents the extraction process adjustment value at position x, and is the adjustment coefficient;

[0138] S524, according to the adjustment function Dynamically adjust key parameters in the digital watermark extraction process, including noise suppression strength, extraction accuracy and digital watermark extraction strategy. The specific extraction strategy includes selecting adaptive filtering algorithm, optimizing digital watermark extraction algorithm, and dynamically adjusting the extraction process in combination with local signal characteristics.

[0139] S525. Utilize the adjusted key parameters of the extraction process to optimize the digital watermark extraction process in real time, optimize its adaptability to different noise interferences and signal changes, and enable the digital watermark to be restored stably and effectively during signal transmission.

[0140] In this implementation manner, the S54 specifically includes:

[0141] S541, based on the local entropy change at position x , construct the mapping relationship between local entropy change and digital watermark recovery accuracy, which quantifies the impact of local entropy change on digital watermark recovery accuracy;

[0142] S542, according to the local entropy change at position x And the signal statistical characteristics of the location , construct a digital watermark recovery optimization function, which is used to dynamically adjust the digital watermark recovery strategy and adaptively adjust the parameters of the digital watermark recovery process according to the comprehensive effect of local entropy changes and signal characteristics:

[0143] ;

[0144] in, represents the optimal value of digital watermark recovery at position x, is a constant offset;

[0145] S543, digital watermark recovery optimization function ,determine the key parameter adjustment strategy in the digital watermark recovery process, and dynamically adjust the parameters in the digital watermark recovery process through real-time feedback of local entropy changes and signal characteristics to achieve efficient recovery optimization;

[0146] S544, restore the optimization function based on the digital watermark The output of the adaptive correction is used to modify the control parameters in the recovery process. This correction mechanism is based on the dynamic response of local entropy changes and signal characteristics, so that the digital watermark recovery process can maintain the optimal recovery accuracy and stability under different signal conditions.

[0147] S545. By optimizing the restoration adjustment parameters, various key operations in the digital watermark restoration process are dynamically corrected, so that the digital watermark information can be restored stably and effectively under the conditions of signal noise interference and signal changes.

[0148] In this implementation manner, S6 specifically includes:

[0149] S61. In a high noise environment, monitor the noise characteristics during signal transmission, and combine the adjusted digital watermark information to optimize the digital watermark embedding process using the inverse information entropy minimization method. By quantifying the impact of signal noise, the noise characteristics are evaluated using the change in local entropy value, and the parameters of the digital watermark embedding process are adjusted to optimize the embedding strategy of the digital watermark information.

[0150] S62, based on the local entropy change at position x And the signal statistical characteristics of the location , construct the optimization objective function in the digital watermark embedding process, and quantify the impact of noise on the accuracy of digital watermark recovery:

[0151] ;

[0152] in, represents the optimization objective function, is the local entropy change at position x, is the noise impact factor at position x, is the adjustment coefficient;

[0153] S63, according to the optimization objective function , dynamically adjust the key parameters in the digital watermark embedding process, and feedback the local entropy changes in real time and noise impact factor , adjust the digital watermark embedding strength, embedding position and digital watermark format to optimize the digital watermark's anti-interference ability;

[0154] S64, based on the optimization objective function The output of is used to adjust the control parameters in the digital watermark embedding process. The adjustment is based on the dynamic response of local entropy changes and signal noise to optimize the digital watermark embedding accuracy and strength, so that the digital watermark maintains the optimal recovery accuracy and stability under different signal conditions.

[0155] S65. In the process of signal transmission, the key parameters of the digital watermark embedding process are dynamically adjusted by combining the inverse information entropy minimization method with the optimized digital watermark embedding strategy. The embedding parameters are adjusted by the output of the signal generator, so that the digital watermark has anti-interference ability during signal transmission and can be effectively restored under noisy conditions.

[0156] Embodiment 1:

[0157] In order to verify the feasibility of the present invention in implementation, the present invention is applied to a digital watermark information protection project carried out by a large communications company. The company is mainly faced with the challenge of embedding and extracting digital watermarks in a highly interfering signal environment, especially in broadcast communication signals. Due to problems such as signal interference, noise and signal attenuation, traditional watermark embedding methods are difficult to ensure the validity and robustness of information. In order to improve the concealment and anti-interference ability of watermarks in signal transmission, the reverse information entropy tracking method of the present invention has been put into practical use.

[0158] The project covers a total area of ​​about 5,000 square kilometers. The main task is to ensure that the digital watermark transmitted through the broadcast signal can be stably and accurately restored at the receiving end. The broadcast signals in this area mainly include communication signals in multiple frequency bands, among which the environmental noise is extremely complex, especially in areas with weak signal strength, where signal noise and interference are more obvious. The traditional method of digital watermark recovery in this area is not ideal, especially in high-noise areas, where the quality of watermark information recovery is seriously affected.

[0159] In order to solve this problem, the present invention first optimizes the digital watermark embedding process based on the high noise environment in the area. Specifically, the present invention dynamically adjusts the watermark embedding process through the inverse information entropy minimization method. By monitoring the change of the local entropy value of the signal in the watermark extraction stage and combining the information entropy adaptive adjustment method, the key parameters in the watermark extraction process can be dynamically optimized. By constructing a watermark recovery optimization function based on the mapping relationship between the local entropy change and the watermark recovery accuracy, the recovery strategy can be adjusted in real time during the watermark recovery process, thereby effectively improving the watermark recovery accuracy.

[0160] During the project implementation, digital watermarks were first embedded in different signal environments in the specified area. Then, actual signal transmission in a high-noise environment was tested, and a comparative experiment was conducted in combination with traditional methods to evaluate the watermark recovery effect. In the actual test, the watermark embedding and recovery process involved the dynamic adjustment of multiple parameters, including entropy value changes, signal statistical feature processing, and watermark recovery optimization functions. The following are the experimental data and comparison results:

[0161] Table 1 Comparison of watermark restoration effects

[0162] ;

[0163] The data in Table 1 show that the present invention has achieved significant improvements in various noise environments, especially in high-noise environments, where the recovery accuracy of the present invention has increased by 30% and the recovery time has decreased by 33%. These data indicate that the present invention can significantly improve the accuracy and efficiency of watermark recovery in complex signal environments, and is particularly suitable for application scenarios with high noise interference such as broadcast communications.

[0164] In order to further verify the stability and adaptability of the present invention in the watermark recovery process, we conducted tests under different signal characteristics. Specifically, we tested the watermark embedding and recovery effects under different signal strengths and different frequency bands. Through the optimized recovery process, the present invention can dynamically adjust to different signal characteristics, further improving the stability of watermark recovery.

[0165] Table 2 Watermark recovery effect under different signal strengths

[0166] ;

[0167] The data in Table 2 show that under different signal strengths, the recovery accuracy of the traditional method is low, especially in a weak signal environment, the watermark recovery effect is almost unfeasible. However, the present invention uses an optimized recovery strategy to not only maintain a high recovery accuracy in a weak signal environment, but also significantly improve the robustness of the watermark recovery process.

[0168] In addition, in the application of the present invention, we also consider the influence of different noise sources, including white noise, narrowband noise and bandwidth-limited signal interference, etc. By combining the inverse information entropy minimization method, the present invention can effectively identify and adapt to different types of noise interference, so that the watermark information can be stably restored in a noisy environment.

[0169] Table 3 Watermark recovery accuracy under different noise types

[0170] ;

[0171] The data in Table 3 show that for different noise types, the watermark recovery accuracy of the present invention is significantly higher than that of the traditional method, especially in white noise and narrowband noise environments, the present invention can provide a more stable and reliable recovery effect.

[0172] In summary, through the implementation of the present invention, the problem of poor robustness and low recovery accuracy of traditional watermark embedding and recovery methods in high noise environments has been successfully solved. The present invention provides higher watermark recovery accuracy and shorter recovery time in high noise and signal change environments through a dynamic adjustment method of reverse information entropy tracking, greatly improving the feasibility and stability of digital watermarks in practical applications. The present invention has broad application prospects in the fields of broadcast communications, data encryption, information protection, etc.

[0173] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A digital watermark camouflage method based on reverse information entropy tracing, characterized in that: The steps include: S1. Monitor the entropy changes of the signal to be embedded, obtain the local entropy value of the signal through the reverse information entropy tracking mechanism, and record the entropy fluctuations of the signal at different processing stages; S2, according to the monitored entropy value change of the signal to be embedded and the local statistical characteristics of the signal, the entropy value gradient is calculated to determine the specific location, timing and strength of the digital watermark embedding; S3, optimize the digital watermark embedding process through the entropy gradient, dynamically adjust the distribution and strength of the digital watermark, so that the digital watermark and signal characteristics can be organically integrated; S4, according to the adjusted digital watermark distribution, monitor the fluctuation of the signal entropy value in real time, use the entropy feedback mechanism to adjust the camouflage degree of the digital watermark, and control the detectability of the digital watermark at different signal processing stages; S5. Based on the entropy value change of the entropy value feedback mechanism, the information entropy adaptive adjustment method is used to dynamically correct the digital watermark extraction and recovery process, so as to achieve effective recovery of digital watermark information under signal changes and noise interference; S6. In a high noise environment, combined with the adjusted digital watermark information, the inverse information entropy minimization method is used to optimize the digital watermark embedding process and the anti-interference ability during signal transmission and processing.

2. According to claim 1, a digital watermark camouflage method based on reverse information entropy tracing is characterized in that: The S1 specifically includes: S11. Apply the reverse information entropy tracking mechanism to calculate the local entropy value of the embedded signal. The local entropy value represents the statistical characteristics of the signal at a specific processing stage and measures the complexity and uncertainty of the signal at this stage. S12, defining a time window for entropy value calculation, wherein the time window corresponds to each processing stage of the signal to be embedded, and in each stage, obtaining a local entropy value change of the stage through real-time sampling and calculation; S13. According to the local statistical characteristics of the signal to be embedded, select a suitable entropy value calculation method, wherein the entropy value calculation method is based on the Shannon entropy formula: ； in, is the local entropy value of signal X in time window t, is the probability distribution of the i-th data point of signal X in time window t; S14, based on the calculated local entropy value , monitor and record the entropy fluctuations of the signal at different processing stages; S15. By analyzing the recorded entropy fluctuations, the signal segment with the smallest entropy fluctuation is identified as the target area for digital watermark embedding, and according to the entropy fluctuations of the signal at each stage, the entropy threshold range is set to perform quantitative analysis.

3. According to claim 1, a digital watermark camouflage method based on reverse information entropy tracing is characterized in that: The S3 specifically includes: S31, based on the calculated entropy value gradient and in combination with the local statistical characteristics of the signal to be embedded, determining the digital watermark embedding strength of different signal regions, and adjusting the digital watermark embedding strength according to the change of the local entropy gradient; S32. Calculate the corresponding digital watermark embedding strength for different regions of the signal through the obtained local entropy gradient: ; in, represents the digital watermark embedding strength at position x, is the digital watermark strength adjustment coefficient, is the absolute value of the local entropy gradient at position x, represents the local statistical characteristics of the signal at position x, Indicates the signal characteristics at that location; S33, based on the calculated digital watermark embedding strength , select the priority embedding area of ​​the signal, the priority embedding area is the area with higher entropy gradient than other areas, and determine the digital watermark embedding position based on the complexity of the area with higher entropy gradient than other areas and the change of entropy gradient; S34, based on the selected priority embedding area and the calculated digital watermark embedding strength , the embedding position and strength of the digital watermark are adjusted, and the digital watermark information and the local characteristics of the signal are weightedly distributed to achieve the fusion of the digital watermark and the signal; S35. Monitor the entropy gradient change during the signal processing process, dynamically optimize the digital watermark embedding process according to the current entropy gradient of the signal, and adjust the digital watermark embedding strategy so that the digital watermark can achieve the desired fusion effect at each processing stage of the signal.

4. According to claim 1, a digital watermark camouflage method based on reverse information entropy tracing is characterized in that: The S4 specifically includes: S41, monitoring and recording the fluctuation of the local entropy value of the signal at different processing stages, extracting the change rate of the local entropy value of each signal segment, and determining the complexity change of the signal based on the change in the entropy value; S42, defining and calculating the entropy value fluctuation threshold of each signal processing stage, and triggering the dynamic adjustment mechanism of the digital watermark camouflage degree when the change rate of the local entropy value exceeds the threshold; S43. According to the monitored local entropy value change rate, a dynamic feedback relationship is established to link the entropy value change with the adjustment function of the digital watermark camouflage degree, which is used to quantify the camouflage degree adjustment: ; in, It represents the change of the local entropy value of the signal at position x between time periods t and t+1. is the adjustment coefficient; S44, based on the feedback entropy value change, by weighted average signal characteristics and entropy value change, calculate the adjustment amount of the digital watermark camouflage degree in each area: ; in, represents the camouflage degree of the digital watermark at position x, is the adjustment weight of the entropy value change on the camouflage degree, is the adjustment weight of signal feature on camouflage degree, Represents the statistical characteristics of the signal at position x; S45. Based on the calculated digital watermark camouflage adjustment value, the digital watermark embedding strategy is dynamically optimized, and the embedding strength and concealment of the digital watermark at different signal processing stages are controlled in real time, so that the detectability of the digital watermark during signal transmission is maintained within the set threshold.

5. According to claim 1, a digital watermark camouflage method based on reverse information entropy tracing is characterized in that: The S5 specifically includes: S51, monitoring the change of the local entropy value of the signal in the digital watermark extraction stage, generating an entropy value feedback signal and performing real-time quantization of the signal characteristics of the digital watermark extraction based on the signal; S52, combining the entropy feedback signal, adjusting the key parameters in the digital watermark extraction process through the information entropy adaptive adjustment method, optimizing the extraction accuracy by using the dynamic characteristics of the entropy value change, and adapting to the influence of signal changes and noise interference; S53, based on the entropy feedback mechanism, a dynamic adjustment strategy based on local entropy changes is adopted to automatically correct the errors in the digital watermark recovery process, so that the accuracy and robustness of the digital watermark recovery process can be improved under the conditions of signal noise interference and signal characteristic changes; S54, constructing a mapping relationship between local entropy change and digital watermark recovery accuracy, and constructing a digital watermark recovery optimization function through the mapping relationship, which adjusts the recovery strategy according to the entropy value change and signal characteristics of the position x: ; in, represents the optimal value of digital watermark recovery at position x, is the adaptive adjustment coefficient, is the local entropy change at position x, Represents the local statistical characteristics of the location, is the signal characteristic adjustment coefficient; S55. Using the recovery adjustment value generated by the digital watermark recovery optimization function, dynamically correct the parameters in the digital watermark extraction and recovery process, so that during the signal transmission process, the digital watermark can be effectively recovered under different noise interference conditions, while ensuring that the concealment and detectability of the digital watermark information remain within a preset threshold.

6. A digital watermark camouflage method based on reverse information entropy tracing according to claim 5, characterized in that: The S52 specifically includes: S521, monitor the change of the local entropy value of the signal in the digital watermark extraction stage, calculate and generate the local entropy change rate of each signal segment , which is used to quantify the dynamic characteristics of the signal in the extraction stage. The entropy change rate is calculated by the difference between the local entropy value of the current signal segment and the local entropy value of the previous signal segment, and then normalized: ; in, represents the local entropy value at position x at time t, Represents the local entropy value at position x at time t-1; S522, based on the local entropy change rate , extract the noise interference characteristics of the signal, and use the noise influence factor To quantify the noise interference of the signal, the noise impact factor is determined by the entropy change rate and the variance of the signal noise: ; in, is the noise impact factor at position x, and is the adjustment coefficient, is the variance of the signal noise at position x; S523, combined with the local entropy change rate and noise impact factor ,Through the information entropy adaptive adjustment method, the key parameters in the digital watermark extraction process are adjusted to optimize the extraction accuracy. The adjustment strategy optimizes the noise suppression and accuracy control in the extraction process according to the dynamic entropy characteristics: ； in, represents the extraction process adjustment value at position x, and is the adjustment coefficient; S524, according to the adjustment function Dynamically adjust key parameters in the digital watermark extraction process, including noise suppression strength, extraction accuracy and digital watermark extraction strategy. The specific extraction strategy includes selecting adaptive filtering algorithm, optimizing digital watermark extraction algorithm, and dynamically adjusting the extraction process in combination with local signal characteristics. S525. Utilize the adjusted key parameters of the extraction process to optimize the digital watermark extraction process in real time, optimize its adaptability to different noise interferences and signal changes, and enable the digital watermark to be restored stably and effectively during signal transmission.

7. According to claim 5, a digital watermark camouflage method based on reverse information entropy tracing is characterized in that: The S54 specifically includes: S541, based on the local entropy change at position x , construct the mapping relationship between local entropy change and digital watermark recovery accuracy, which quantifies the impact of local entropy change on digital watermark recovery accuracy; S542, according to the local entropy change at position x And the signal statistical characteristics of the location , construct a digital watermark recovery optimization function, which is used to dynamically adjust the digital watermark recovery strategy and adaptively adjust the parameters of the digital watermark recovery process according to the comprehensive effect of local entropy changes and signal characteristics: ; in, represents the optimal value of digital watermark recovery at position x, is a constant offset; S543, digital watermark recovery optimization function ,determine the key parameter adjustment strategy in the digital watermark recovery process, and dynamically adjust the parameters in the digital watermark recovery process through real-time feedback of local entropy changes and signal characteristics to achieve efficient recovery optimization; S544, restore the optimization function based on the digital watermark The output of the adaptive correction is used to modify the control parameters in the recovery process. This correction mechanism is based on the dynamic response of local entropy changes and signal characteristics, so that the digital watermark recovery process can maintain the optimal recovery accuracy and stability under different signal conditions. S545. By optimizing the restoration adjustment parameters, various key operations in the digital watermark restoration process are dynamically corrected, so that the digital watermark information can be restored stably and effectively under the conditions of signal noise interference and signal changes.

8. According to claim 1, a digital watermark camouflage method based on reverse information entropy tracing is characterized in that: The S6 specifically includes: S61. In a high noise environment, monitor the noise characteristics during signal transmission, and combine the adjusted digital watermark information to optimize the digital watermark embedding process using the inverse information entropy minimization method. By quantifying the impact of signal noise, the noise characteristics are evaluated using the change in local entropy value, and the parameters of the digital watermark embedding process are adjusted to optimize the embedding strategy of the digital watermark information. S62, based on the local entropy change at position x And the signal statistical characteristics of the location , construct the optimization objective function in the digital watermark embedding process, and quantify the impact of noise on the accuracy of digital watermark recovery: ; in, represents the optimization objective function, is the local entropy change at position x, is the noise impact factor at position x, is the adjustment coefficient; S63, according to the optimization objective function , dynamically adjust the key parameters in the digital watermark embedding process, and feedback the local entropy changes in real time and noise impact factor , adjust the digital watermark embedding strength, embedding position and digital watermark format to optimize the digital watermark's anti-interference ability; S64, based on the optimization objective function The output of is used to adjust the control parameters in the digital watermark embedding process. The adjustment is based on the dynamic response of local entropy changes and signal noise to optimize the digital watermark embedding accuracy and strength, so that the digital watermark maintains the optimal recovery accuracy and stability under different signal conditions. S65. In the process of signal transmission, the key parameters of the digital watermark embedding process are dynamically adjusted by combining the inverse information entropy minimization method with the optimized digital watermark embedding strategy. The embedding parameters are adjusted by the output of the signal generator, so that the digital watermark has anti-interference ability during signal transmission and can be effectively restored under noisy conditions.

Images