Method for copyright authentication of online course based on time domain hidden watermark

Abstract

The application provides a network course copyright authentication method based on a time domain hidden watermark, and relates to the technical field of electric digital data processing.The micro noise signal in a display screen driving circuit is collected from a client, time domain sampling, filtering normalization and principal component analysis dimension reduction processing of the noise signal are carried out, a unique noise feature vector NS is extracted, and the noise feature vector NS is further combined with local video statistical parameters of the video frame of the client, so that the dynamic adjustment of the watermark embedding amplitude and position is realized by using an adaptive modulation model.The randomness and non-reproducibility of the noise feature are utilized, and the unique generation of the watermark key is realized through the coupling of the local video parameters and the nonlinear mapping function, so that the security and the tamper resistance of the copyright authentication are significantly improved, and a feasible technical path for realizing the unique determination of the watermark key is provided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital data processing, in particular to a network course copyright authentication method based on time domain hidden watermark. BACKGROUND

[0002] Traditional digital watermarking schemes mostly rely on frequency domain, spatial domain or hybrid domain methods, and realize the hiding and transmission of copyright information through transform domain decomposition, feature extraction and watermark modulation processes. Among them, the frequency domain watermarking technology uses discrete cosine transform (DCT), discrete wavelet transform (DWT) and other methods to embed the watermark into the medium and low frequency coefficients, so as to maintain the robustness of the watermark information in the compression and transmission process; while the spatial domain watermarking technology directly operates on the pixel value, and has the advantages of simple implementation and high real-time performance. However, the traditional technology generally faces the limitations of insufficient anti-interference of embedded watermark, easy tampering of embedding position, and influence on the stability of watermark signal in high noise environment in practical application.

[0003] In the prior art, a method for adding invisible watermark to client is disclosed in CN110543749A, which obtains the invisible watermark processing request sent by the user, the user information and the pure color picture selected by the user, transforms the user information according to the invisible watermark processing request, generates a unique first watermark ID, stores the first watermark ID in the watermark database, then synthesizes the first watermark ID and the pure color picture to generate an invisible watermark picture with the first watermark ID, and finally returns the invisible watermark picture to the user for the user to use the invisible watermark picture as the background of the client;

[0004] However, there are still many deficiencies in practical application, especially in the multi-scene application of network course, which requires higher copyright information uniqueness, robustness and stability; the existing technology mostly relies on traditional frequency domain or spatial domain watermarking methods, which often leads to watermark signal loss or distortion due to local suppression or reconstruction error when facing complex processing methods such as intelligent video compression, editing and malicious tampering; at the same time, the traditional scheme fails to fully utilize the microscopic noise information in the display screen driving circuit and the local video statistical parameters of the client, which have potential advantages in ensuring the unpredictability and uniqueness of embedding;

[0005] The above information disclosed in the above background section is only used to strengthen the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0006] The present application aims to provide a network course copyright authentication method based on time domain hidden watermark to solve the problems raised in the above background.

[0007] In order to achieve the above object, the present application provides the following technical scheme:

[0008] The online course copyright authentication method based on the time domain hidden watermark specifically comprises the following steps:

[0009] Step S1: collecting micro noise signals in a display screen driving circuit in a client, wherein the micro noise signals are digital noise data sequences obtained by continuous sampling in a time domain sampling window;

[0010] Step S2: filtering and normalizing the collected digital noise data sequences according to a unified time domain block, and extracting a preliminary noise statistical feature sequence of the digital noise data sequences;

[0011] Step S3: performing dimension reduction on the extracted preliminary noise statistical feature sequence by using a principal component analysis method, to obtain a noise feature vector NS;

[0012] Step S4: dynamically modulating the amplitude and position of the time domain watermark embedding by using an adaptive modulation model based on local video statistical parameters of video frames in the client audio and video data, wherein the modulation process utilizes a nonlinear function to couple the local video statistical parameters and the watermark embedding amplitude;

[0013] Step S5: combining the noise feature vector NS and the adaptive modulation model, and generating a unique watermark key by using a determined nonlinear mapping function;

[0014] Step S6: coupling the watermark key and watermark modulation parameters in the time domain hidden watermark embedding process, wherein the coupling process involves adjusting the watermark embedding position and amplitude in a time domain data window of the audio and video data.

[0015] Compared with the prior art, the present application has the following beneficial effects: by collecting micro noise signals in a display screen driving circuit in a client, performing time domain sampling, filtering and normalizing on the noise signals, and performing principal component analysis dimension reduction processing, a unique noise feature vector NS is extracted, and further combined with local video statistical parameters of video frames in the client, an adaptive modulation model is used to realize dynamic adjustment of the watermark embedding amplitude and position; not only the randomness and non-replicability of the noise feature are utilized, but also the unique generation of the watermark key is realized through the coupling of the local video parameters and the nonlinear mapping function, which significantly improves the security and tamper resistance of the copyright authentication; a practical technical path for realizing unique determination of the watermark key is provided. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 The present application is a whole method flowchart. DETAILED DESCRIPTION

[0017] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with specific examples.

[0018] It should be noted that, unless otherwise defined, technical terms or scientific terms used in the present application should be understood as their common meanings to those skilled in the art to which the present application pertains. The terms "first", "second", and similar terms used in the present application do not represent any order, number, or importance, but are only used to distinguish different components. The terms "include" or "contain" and similar terms mean that the elements or objects before the terms encompass the elements or objects listed after the terms and their equivalents, without excluding other elements or objects. The terms "connect" or "connected" and similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The terms "upper", "lower", "left", "right", and the like only represent relative positional relationships, which can change accordingly when the absolute positions of the described objects change.

[0019] Example 1

[0020] Please refer to Figure 1 The present application provides a technical solution:

[0021] The online course copyright authentication method based on time domain hidden watermark includes the following specific steps:

[0022] Step S1: Collecting micro-noise signals in the display screen driving circuit on the client side, wherein the micro-noise signals are digital noise data sequences obtained by continuous sampling in the time domain sampling window.

[0023] Further description: The specific operation steps are as follows:

[0024] The collection of micro-noise signals is realized by the cooperation of the high-precision analog-to-digital converter (ADC) and the low-noise amplifier of the collection module.

[0025] In the client device, a set of integrated high-precision ADC circuit and low-noise amplifier are arranged and connected to the test point of the display screen driving circuit to collect the micro-noise signals existing in the circuit in real time, wherein the micro-noise signals include thermal noise signals and random current noise signals.

[0026] Thermal noise signal: Originating from the thermal motion of electrons in the conductor to which the client belongs, it can cause random fluctuations in voltage and current. This noise changes with temperature and is calculated by the root mean square value formula, which has a wide frequency characteristic.

[0027] Characterization of thermal noise signal:

[0028] Root mean square voltage (RMS Voltage):

[0029] Thermal noise is usually characterized by voltage, whose root mean square voltage can be calculated by the formula:

[0030] ;

[0031] wherein, is the root mean square value of thermal noise voltage; is the Boltzmann constant, taking the value of ; T is the absolute temperature (unit: Kelvin); R is the resistance value (unit: Ohm); B is the bandwidth (unit: Hertz).

[0032] Spectral characteristics: Thermal noise has white noise characteristics, that is, it is uniformly distributed throughout the frequency spectrum, that is, its power spectral density is equal at all frequencies.

[0033] Random current noise signal: It is obvious at low frequencies, caused by random changes in current and material inhomogeneity. This noise is also called flicker noise or "1 / f noise", whose intensity is related to frequency;

[0034] Characterization of random current noise signal:

[0035] Power spectral density (Power Spectral Density, PSD):

[0036] Random current noise is usually significant in the low frequency region, and its power spectral density can be defined as:

[0037] ;

[0038] wherein, is the power spectral density of noise current; is a constant related to material and process; is the frequency; is the spectral index, taking a value between 0.5 and 1.5.

[0039] 1 / f noise characteristics:

[0040] Random current noise is "flicker noise" or "1 / f noise", which is characterized by high power spectrum in the low frequency band and decreases with increasing frequency.

[0041] Record the original analog signal of the collected data and transmit it to the subsequent digitizing unit;

[0042] According to the fixed time domain sampling window, the collected microscopic noise signal is digitized to obtain a continuous digital noise data sequence; specifically: the fixed time domain sampling window of the embodiment is a sampling period of 50 milliseconds; in each 50 millisecond sampling period, the analog microscopic noise signal output from the acquisition module is input to the ADC module;

[0043] The ADC module converts the data in each time domain sampling window according to the set sampling rate, and outputs a continuous digital noise data sequence.

[0044] The buffer stores the digital noise data sequence in each time domain sampling window obtained by continuous sampling, for subsequent preprocessing.

[0045] The digital noise data sequence in each time domain sampling window output from the ADC module is stored in the local buffer system.

[0046] The buffer design adopts a circular storage structure, so that the continuous sampling data sequence can be continuously saved.

[0047] After the data buffer is completed, it provides a data source for subsequent preprocessing and watermark embedding.

[0048] Step S2: After filtering and normalizing the collected digital noise data sequence according to a unified time domain block, the preliminary noise statistical feature sequence of the digital noise data sequence is extracted.

[0049] Further explanation: The digital noise data sequence in each time domain sampling window buffered is filtered by a digital low-pass filter to eliminate high-frequency interference clutter.

[0050] Specifically, the digital low-pass filtering process is performed on the digital noise data sequence in each time domain sampling window stored in the client buffer.

[0051] The digital low-pass filter parameters are configured, and the cutoff frequency is set to ensure that the high-frequency interference clutter is effectively suppressed from the data sequence.

[0052] The filtered data sequence is output to the next processing unit.

[0053] The filtered digital noise data sequence is normalized, and the processing result falls within the preset numerical interval [0, 1].

[0054] In each time domain sampling window, the digital noise data sequence is calculated by combining the edge detection algorithm and the moving average method to generate a preliminary noise statistical feature sequence; the statistical features include but are not limited to mean, variance, and extreme value; specifically:

[0055] Using the normalized digital noise data sequence, first, the first-order differential detection of the edge detection algorithm is implemented to identify the edge information of the value fluctuation in the sequence.

[0056] On the basis of the edge detection result, the sliding average method is used, and the fixed window size is set to 10 sampling points to smooth the data sequence.

[0057] Calculate the statistical parameters of the smoothed data sequence respectively, including the mean, variance, and the extreme value represented by the maximum or minimum value in the sequence, to form a preliminary noise statistical feature sequence;

[0058] Pass the generated preliminary noise statistical feature sequence to the subsequent step S3.

[0059] Step S3: Perform dimension reduction on the extracted preliminary noise statistical feature sequence using principal component analysis to obtain a noise feature vector NS;

[0060] Further explanation: Convert the preliminary noise statistical feature sequence into a noise feature vector of fixed dimension; the specific operation steps are as follows:

[0061] Take the normalized preliminary noise statistical feature sequence in each time domain sampling window as an input data matrix, with each row of the data matrix corresponding to a different time domain sampling window and each column being a normalized statistical indicator;

[0062] Perform covariance matrix calculation on the input data matrix using the standard PCA algorithm to obtain eigenvalues and eigenvectors; specifically:

[0063] Perform covariance matrix calculation, the formula of which is , where X is the input data matrix, and μ is the mean of each column;

[0064] Perform eigenvalue and eigenvector solving operation of the covariance matrix to obtain all eigenvalues and corresponding eigenvectors;

[0065] Arrange the obtained eigenvalues in descending order and arrange the eigenvectors correspondingly;

[0066] Select the first k principal components so that the cumulative contribution rate reaches the preset contribution rate threshold; the preset contribution rate threshold is set to 90% in this embodiment;

[0067] Specifically, construct the selected principal components into a noise feature vector NS of fixed length, i.e. ; , where represents the noise feature component of the kth principal component; specifically:

[0068] Calculate the contribution rate of each principal component from the obtained eigenvalues arranged in descending order, and the single contribution rate is: ;

[0069] Cumulative contribution rate is sequentially added , until the cumulative value is greater than or equal to the preset contribution rate threshold of 90%, and the first k principal components are determined, where k is the smallest integer that satisfies i =1 k CR i ≥90%.

[0070] Σ i =1 k CR i ≥90% represents the cumulative contribution rate of the first item i=1, and then to the kth item, until the cumulative contribution rate reaches or exceeds the preset contribution rate threshold 90%;

[0071] The noise feature vector NS is constructed as , wherein represents the feature component corresponding to the ith principal component.

[0072] The constructed noise feature vector NS is serialized and stored in the order of each time domain sampling window; ensure that the noise feature vector forms a stable input in the time domain order, the specific implementation steps are:

[0073] According to the constructed noise feature vector NS, the corresponding time domain sampling window in the original data matrix is numbered in the order of arrangement;

[0074] Use the serialization module to store each NS in a fixed format, and ensure that the storage format is uniform;

[0075] Ensure that the NS sequence after serialized storage forms a continuous and stable input in the time domain order, providing data basis for subsequent watermark embedding and copyright authentication.

[0076] Step S4: Based on the local video statistical parameters of the video frames in the client audio and video data, the amplitude and position of the time domain watermark embedding are dynamically modulated through an adaptive modulation model, and the modulation process utilizes a nonlinear function to couple the local video statistical parameters and the watermark embedding amplitude;

[0077] Further explanation: In the audio and video data, the continuous video frame sequence is divided according to the fixed time domain sampling window, and the video frame image corresponding to each time domain sampling window is extracted;

[0078] Divide the video frame image into multiple regions, denoted as {1, 2, …, r, …, R}, where r represents the rth divided region in the video frame image, and R is the total number of divided regions;

[0079] Perform local light intensity analysis on each divided region of the video frame image in each time domain sampling window, and calculate the local video statistical parameters of each divided region, including the local light intensity mean , and the variance , and the motion vector amplitude ;

[0080] It should be noted that the light intensity mean of all pixel points in region r is calculated using an image processing algorithm, and is recorded as The variance of the light intensity values of all the pixel points in the region r is calculated, and recorded as ;

[0081] The optical flow of the pixel motion between adjacent video frame images is calculated to obtain the motion vector of each region r;

[0082] The motion vector amplitude of the region r is calculated and recorded as ;

[0083] The motion vector amplitude of each region in each time domain sampling window is extracted by processing the pixel motion between consecutive video frame images using the optical flow method;

[0084] The light intensity change region of the video frame image is detected, and the light intensity change region is divided into a light intensity change significant region and a light intensity change smooth region by setting a light intensity gradient edge detection threshold;

[0085] It should be noted that: for each divided region r, the Sobel operator of the edge detection algorithm is used to calculate the light intensity gradient; a preset light intensity gradient threshold ;

[0086] The region with a light intensity change greater than or equal to the light intensity gradient threshold is divided into a light intensity change significant region, and the rest is divided into a light intensity change smooth region;

[0087] The light intensity change classification results of each region are recorded and marked as a significant region set and a smooth region set, respectively;

[0088] The adaptive modulation model is defined to include:

[0089] The region with a motion vector amplitude lower than a preset threshold is marked, and is compared with the light intensity change region, and the region with a light intensity change smooth and a motion vector amplitude lower than a preset threshold is preferentially selected as a target position for watermark embedding, and these regions are recorded as a target position region set, denoted as {1, 2, …, r1, …, R1}, wherein r1 represents the index of the target position for watermark embedding, R1 is the total number of target positions for watermark embedding, and r1∈{1, 2, …, R}, R1≤R;

[0090] The preset threshold of the motion vector amplitude is set to ;

[0091] The region with a motion vector amplitude higher than the preset threshold is dynamically excluded to ensure that the watermark is not lost or has artifacts after being embedded due to compression or clipping.

[0092] Using the mean, variance, and motion vector amplitude of the local light intensity calculated from each region of the target location area set as input, a coupling function between the watermark embedding amplitude and these parameters is constructed using a nonlinear least squares regression method, forming the following adaptive modulation model of the target location r1 with respect to the watermark embedding amplitude:

[0093] ;

[0094] in, The initial watermark embedding amplitude at target position r1 modulated under time domain variable t. The base amplitude of the watermark at target position r1; This is a nonlinear function relating the mean and variance of local light intensity at target location r1; in this embodiment... It is a quadratic polynomial or exponential model. The function is a nonlinear function related to the magnitude of the motion vector at the target position r1, where α and β are coupling coefficients; it was obtained through fitting based on field experiments; this embodiment The nonlinear function is represented by the log or sigmoid function.

[0095] It should be noted that:

[0096] definition Using a quadratic polynomial form, the specific formula is as follows:

[0097] ;

[0098] in: The mean local light intensity at target location r1;

[0099] Let be the local light intensity variance at target location r1;

[0100] , c1 and c2 are regression coefficients obtained by fitting field experimental data using a nonlinear least squares regression method;

[0101] definition The Sigmoid function is used, and the specific formula is as follows:

[0102] ;

[0103] in, The magnitude of the motion vector at the target position r1; These are the regression coefficients obtained by fitting field experimental data using a nonlinear least squares regression method.

[0104] Determine the regression coefficients , c1 and The specific steps are as follows:

[0105] Collecting field experiment data, including the values of 、 and of multiple target positions and the corresponding ideal watermark embedding amplitudes ;

[0106] Constructing a least squares regression model, and estimating the regression coefficients 、 , c1 and by the gradient descent method of the optimization algorithm, so that the fitting functions f and g can best match the experimental data and minimize the prediction error.

[0107] The following test content is given in this embodiment:

[0108] The experiment takes multiple client audio and video data as the main test object, selects video materials under different dynamic scenes for analysis and testing, in order to verify the innovation and advantages of the time domain watermark embedding mechanism based on local video statistical parameters; the selected video materials include "city street night view video", "indoor static interview video", "sports event video", etc., covering high dynamic and low dynamic scenes. The experiment uses the coupling mechanism of optical flow detection, edge detection and adaptive modulation model to analyze the watermark embedding behavior, and records the related data. The specific steps and implementation process are as follows:

[0109] Video sampling and division:

[0110] The content of the experimental video sample is divided into continuous video frame sequences according to a 50 millisecond time domain sampling window, ensuring that the sampling time domain window has sufficient precision and stability.

[0111] Each video frame image is divided into a 5x5 area grid by uniform division processing, obtaining 25 areas, each of which is marked as 1 to 25. These areas provide a basis for subsequent local statistical parameters and target position area selection.

[0112] Local statistical parameter extraction:

[0113] Light intensity analysis: calculate the local light intensity value of each pixel in the region, and extract the light intensity mean and variance. The light intensity mean describes the overall brightness level of the region, and the variance reflects the fluctuation range of the pixel value.

[0114] Motion vector calculation: calculate the motion vector between consecutive frames by the optical flow method, and extract the motion amplitude of each region as a dynamic change indicator.

[0115] In the experimental design, the light intensity mean, variance and motion vector data are effectively recorded and normalized to the numerical interval [0, 1].

[0116] Gradient edge detection is applied, and a light flow method is used to analyze the change of local motion vectors in different video scenes.

[0117] The region with smooth light intensity change and motion vector amplitude lower than a preset threshold (0.4) is selected as a target position region for watermark embedding, and the salient region is excluded to ensure storage stability.

[0118] Adaptive modulation model construction and watermark embedding:

[0119] Based on the mean, variance and motion vector amplitude of the target position region, a dynamic watermark modulation model is established through a nonlinear coupling function.

[0120] The watermark embedding amplitude and position of the target position region are modulated in real time, and the sequence storage is adopted to form a stable and verifiable watermark.

[0121] The amplitude modulation range of the data containing the embedded watermark, the integrity of the video image after watermark generation, and the watermark preservation after compression and clipping.

[0122] Different dynamic scenes are called for testing, and the embedding efficiency, error rate and watermark robustness are recorded.

[0123] The following table records the local statistical parameters of the experimental video and the verification results after watermark embedding:

[0124] Table 1 Innovation and advantage of time domain watermark embedding mechanism of local video statistical parameters:

[0125]

[0126] The experimental results show that:

[0127] The dynamic modulation mechanism of the watermark embedding amplitude is based on the local statistical parameters of the video frame, which can adaptively adjust the embedding range for different dynamic scenes, effectively improving the robustness.

[0128] The embedded watermark maintains high integrity after compression and clipping, verifying the stability of the adaptive modulation model in the invention.

[0129] Compared with the traditional watermark embedding method based on the overall parameters of the frame, the method selects the target position region by local light intensity and motion vector, effectively reduces the embedding position deviation and information loss risk, and reflects high innovation.

[0130] Step S5: Combine the noise feature vector NS with the adaptive modulation model to generate a unique watermark key through a determined nonlinear mapping function;

[0131] Further description: the nonlinear mapping function is based on the time domain information to nonlinearly couple the noise features and the local video statistical parameters, to generate a fixed format watermark key; the watermark key is denoted as ; the specific implementation steps include:

[0132] The following nonlinear mapping function is predetermined:

[0133]

[0134] wherein, is each component in the noise feature vector NS, and δ are constant parameters obtained by fitting offline experimental data; is a fusion coefficient;

[0135] The generated noise feature vector NS and the calculated illumination and motion parameters are sequentially substituted into the nonlinear mapping function, to calculate a unique and fixed length watermark key ;

[0136] The watermark key is combined with the time domain related timestamp data, to ensure the stability and uniqueness of the generated key in the entire time domain;

[0137] The watermark key is stored in the watermark key module; and the watermark key module is used as a key parameter input of the subsequent time domain watermark embedding algorithm;

[0138] In this embodiment, the determination manner of the nonlinear mapping function is as follows:

[0139] Through offline experimental data fitting, the constant parameters , δ and the fusion coefficient in the nonlinear mapping function are determined.

[0140] An experimental data set containing multiple groups of noise feature vectors NS and corresponding local video statistical parameters is collected, including the , , and values of multiple target position regions;

[0141] Using a nonlinear least squares regression method, based on the collected experimental data, the constant parameters and δ in the nonlinear mapping function are fitted and estimated, and the specific steps are as follows:

[0142] An objective function is constructed, which is defined as the sum of squares of fitting errors;

[0143] Gradient descent method or other optimization algorithms are used to iteratively adjust the parameters , δ and the fusion coefficient Continue until the objective function reaches its minimum value;

[0144] Verify the fitting results to ensure that the parameters remain consistent across different target locations, thus guaranteeing the uniformity and stability of the nonlinear mapping function.

[0145] Step S6: During the time-domain hidden watermark embedding process, the watermark key is coupled with the watermark modulation parameters. The coupling process involves adjusting the watermark embedding position and amplitude within the time-domain data window of the audio and video data.

[0146] Further explanation: The audio and video data includes online course videos or audio data. Within the online course videos or audio data, continuous time-domain data windows are divided. In this embodiment, each time-domain data window corresponds to audio and video data within 10 milliseconds.

[0147] The watermark modulation parameters include the initial watermark embedding amplitude. and time-domain variable t;

[0148] The generated watermark key As a dynamic seed, it is related to the initial watermark embedding amplitude. Temporal modulation calculations are performed using local video statistical parameters to determine the final watermark embedding amplitude and final embedding position within each temporal data window;

[0149] Explanation of the final watermark embedding range:

[0150] Within each time-domain data window, according to the modulation function Preliminary watermark embedding range Perform dynamic calculations, where For watermark-based keys The nonlinear modulation function with respect to the time-domain variable t; It is the final watermark embedding amplitude of the target location r1 region;

[0151] Explanation of the final embedding location:

[0152] Within each time domain data window, combined with and Perform the following calculations based on the modulation function:

[0153] ;

[0154] in, It is the location filtering factor for the target location region r1; it filters out regions from the target location region set {1,2,...,r1,...,R1}. The region with the smallest value is used as the final embedding location;

[0155] Specifically: from the target position area set {1, 2,..., r1,..., R1}, obtain the position filtering factor of all areas r1 ;

[0156] Compare the values of each area Select the area with the smallest value as the final watermark embedding position ;

[0157] If there are multiple areas The value is the same and the smallest, select the area with the smallest index in the matrix as the final embedding position

[0158] Ensure that only one final embedding position is selected in each time domain data window, ensuring the consistency and uniqueness of watermark embedding.

[0159] It should be noted that The smaller the value, the smaller the value And / or The smaller the corresponding local light intensity mean and variance, the more stable the pixel brightness of the corresponding area, which is beneficial to the selection of watermark embedding position

[0160] Determine the mathematical expression of the nonlinear modulation function The specific formula is as follows:

[0161] ;

[0162] Wherein, Is the unique watermark key generated under the time domain variable t; t is the time variable of the current time domain data window, in milliseconds

[0163] For each time domain data window t, perform the following operations:

[0164] Obtain the final watermark embedding amplitude ; Obtain the final embedding position r2, r2∈{1, 2,..., R1}

[0165] In the time domain data window t of the original audio and video data, for the area r2, embed the watermark signal S(t) according to the following formula:

[0166] ;

[0167] Wherein, Is the signal of the original audio and video data in the time domain data window t

[0168] S(t) is a pre-defined watermark signal Is the audio and video data after embedding the watermark

[0169] Ensure that the embedding process is only performed for the final embedding position r2 within each time domain data window, avoiding multiple embedding leading to signal distortion.

[0170] The final watermark embedding amplitude after modulation And the final embedding position is embedded in the original audio and video data in a stealthy way, and the watermark embedding result forms the final watermark embedding video; and the watermark embedding video is sent to the client through the data transmission interface for online course copyright authentication.

[0171] It should be noted that all the calculation formulas in the present application file use regression analysis including but not limited to machine learning algorithms to deeply analyze the collected relevant parameters and identify their natural trend and mutual relationship. Professional software such as Python's Scikit-learn library or R language is used to automatically generate mathematical models matching the data. Then, the performance of the model is objectively evaluated through cross-validation and other methods, and combined with continuous feedback and optimization to ensure that the created formula truly reflects the inherent law of the data, thereby ensuring its effectiveness and accuracy. In all the calculation formulas in the present application, the parameters in each formula are processed by consistent range of dimensionless to ensure that different physical quantities are compared on the same scale; the dimensionless technology means include but are not limited to Min-Max Normalization, Z-Score standardization;

[0172] The technical solutions of the present application can be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), FLASH, hard disk or optical disk, etc., including a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute the method of each embodiment of the present application.

[0173] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered a list of executable instructions for implementing logic functions, and can be specifically embodied in any computer-readable medium for use by an instruction execution system, apparatus or device, such as a computer-based system, a system including a processor, or other system that can take instructions from an instruction execution system, apparatus or device, or in conjunction with these instruction execution systems, apparatus or devices. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport programs for use by an instruction execution system, apparatus or device, or in conjunction with these instruction execution systems, apparatus or devices.

[0174] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not to limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced without departing from the spirit and scope of the present application, and all modifications and equivalents should be included in the scope of the claims of the present application.

Claims

1. A method for online course copyright authentication based on time domain hidden watermark, characterized in that, The specific steps include: Step S1: Collecting micro-noise signals in the display screen driving circuit of the client, the micro-noise signals including thermal noise signals and random current noise signals; digitizing the collected micro-noise signals according to a fixed time domain sampling window to obtain a continuous digital noise data sequence; Thermal noise signals: resulting from the thermal motion of electrons in the conductor to which the client belongs, which can cause random fluctuations in voltage and current; Random current noise signals: obvious at low frequencies, caused by random changes in current and material inhomogeneity; Step S2: filtering and normalizing the collected digital noise data sequence according to a unified time domain block to extract a preliminary noise statistical feature sequence of the digital noise data sequence; Step S3: using principal component analysis to reduce the dimension of the extracted preliminary noise statistical feature sequence to obtain a noise feature vector NS; Step S4: based on the local video statistical parameters of the video frames in the audio and video data of the client, dynamically modulating the amplitude and position of the time domain watermark embedding through an adaptive modulation model, the modulation process using a nonlinear function to couple the local video statistical parameters and the watermark embedding amplitude; Step S5: combining the noise feature vector NS with the adaptive modulation model to generate a unique watermark key through a determined nonlinear mapping function; Step S6: in the process of time domain hidden watermark embedding, coupling the watermark key with the watermark modulation parameters, the coupling process involving adjusting the watermark embedding position and amplitude in the time domain data window of the audio and video data; forming the final watermark embedded video after embedding the watermark, and sending the watermark embedded video to the client through the data transmission interface for online course copyright authentication.

2. The time-domain hidden watermark-based online course copyright authentication method according to claim 1, characterized in that: The cache obtains the digital noise data sequence in each time domain sampling window obtained by continuous sampling.

3. The time-domain hidden watermark-based online course copyright authentication method according to claim 2, characterized in that: The digital low-pass filter is used to filter the digital noise data sequence in each time domain sampling window cached to eliminate high-frequency interference clutter; The filtered digital noise data sequence is normalized, and the processing result falls within the preset numerical interval [0, 1]; In each time domain sampling window, the digital noise data sequence is calculated by combining the edge detection algorithm with the sliding average method to generate a preliminary noise statistical feature sequence; the statistical features include but are not limited to mean, variance, and extreme value.

4. The time-domain hidden watermark-based online course copyright authentication method according to claim 3, characterized in that: The preliminary noise statistical feature sequence is converted into a noise feature vector of a fixed dimension. The selected principal components are specifically constructed into a fixed-length noise feature vector NS, that is ; denotes the noise feature component of the kth principal component. The constructed noise feature vector NS is serialized and stored in the order of each time domain sampling window.

5. The time-domain hidden watermark-based online course copyright authentication method according to claim 4, characterized in that: In the audio and video data, the continuous video frame sequence is divided according to the fixed time domain sampling window, and the video frame image corresponding to each time domain sampling window is extracted; The video frame image is divided into multiple regions, denoted as {1, 2, …, r, …, R}, where r represents the rth divided region in the video frame image, and R is the total number of divided regions; Local light intensity analysis is performed on each divided area of the video frame image in each time domain sampling window, and local video statistical parameters of each divided area are calculated, wherein the local video statistical parameters include local light intensity mean value and variance and motion vector amplitude ; The optical flow method is used to process the pixel motion between consecutive video frame images to extract the motion vector amplitude of each region in each time domain sampling window; The light intensity change region of the video frame image is detected, the light intensity change region is divided into a light intensity change significant region and a light intensity change smooth region by setting a light intensity gradient edge detection threshold.

6. The time-domain hidden watermark-based online course copyright authentication method according to claim 5, characterized in that: The adaptive modulation model is defined, and includes: Regions of the pixel region between video frame images, whose motion vector amplitudes are lower than a preset threshold, are marked and compared with the light intensity change region, regions of the light intensity change smooth region and the motion vector amplitude lower than the preset threshold are selected as target positions for watermark embedding, and these regions are recorded as a target position region set, recorded as {1, 2, …, r1, …, R1}, wherein r1 represents a target position index for watermark embedding, R1 is the total number of target positions for watermark embedding, and r1 ∈ {1, 2, …, R}, R1 ≤ R; Regions of the pixel region between video frame images, whose motion vector amplitudes are higher than a preset threshold, are dynamically excluded; The local light intensity mean value, variance and motion vector amplitude of each region in the target position region set are taken as inputs, a coupling function of the watermark embedding amplitude and these parameters is constructed by using a nonlinear least squares regression method, and the following adaptive modulation model of the target position r1 with respect to the watermark embedding amplitude is formed: ；  wherein, is a preliminary watermark embedding amplitude modulated by the target position r1 at the time domain variable t, is a watermark base amplitude of the target position r1; is a nonlinear function related to the local light intensity mean and variance at the target position r1; is a nonlinear function related to the motion vector amplitude at the target position r1, and α and β are coupling coefficients.

7. The time-domain hidden watermark-based online course copyright authentication method according to claim 6, characterized in that: The nonlinear mapping function performs nonlinear coupling on noise features and local video statistical parameters based on time domain information, and generates a fixed format watermark key; The generated noise feature vector NS is substituted into the nonlinear mapping function and The nonlinear mapping function is substituted in turn, and the unique and fixed-length watermark key is calculated .

8. The time-domain hidden watermark-based online course copyright authentication method according to claim 7, characterized in that: The audio and video data includes online course video or audio data, and continuous time domain data windows are divided in the online course video or audio data; The watermark modulation parameters include a preliminary watermark embedding amplitude and a time domain variable t; The generated watermark key As a dynamic seed, respectively with the preliminary watermark embedding amplitude And local video statistical parameters are modulated in time domain to calculate the final watermark embedding amplitude and the final embedding position in each time domain data window. The following is a description of the final watermark embedding amplitude: Within each time domain data window, a modulation function is set The preliminary watermark embedding amplitude is dynamically calculated, wherein is based on a watermark key and a non-linear modulation function of the time domain variable t. is the final watermark embedding amplitude for the target location r1 region. The following is a description of the final embedding position: Within each time-domain data window, combine and The following calculations are made in accordance with the modulation function: ； Wherein, is the position screening factor of the target position r1 area; screening from the target position area set {1, 2, …, r1, …, R1} The area with the smallest value is the final embedding position; Final watermark embedding amplitude after modulation of the signal to embed the watermark and the final embedding position are embedded in the original audiovisual data in a stealthy manner.

Images