Blockchain-based copyright data protection method, system, device and storage medium

Abstract

This invention discloses a blockchain-based method, system, device, and storage medium for copyright data protection. The method includes: performing multi-layer compression sensing processing on the original copyright data to obtain a feature watermark; fusing the feature watermark and a real watermark to obtain a target watermark, wherein the real watermark and the feature watermark have the same size; uploading the target watermark to a distributed storage system and uploading the hash value corresponding to the target watermark to the blockchain. This improves both the robustness and security of the watermark.

Description

Technical Field

[0001] This invention relates to the field of blockchain technology, and in particular to a blockchain-based method, system, device, and storage medium for copyright data protection. Background Technology

[0002] With increasing digitalization and informatization, copyrighted data, such as images, audio, and video, are being used more and more widely on social networks as carriers of information. During this widespread dissemination, there is a significant risk of copyrighted data theft, making copyright data protection a growing concern. Digital watermarking technology is an important technology for copyright data protection. Blockchain technology can be used to achieve even more secure and reliable digital watermarking.

[0003] In related technologies, the discrete cosine transform method is used to add digital watermarks to copyright data to protect it. However, this method embeds the watermark directly into the original copyright data, requiring modification of the original data, which leads to distortion and poor robustness of the original copyright data. Furthermore, the added digital watermark is easily tampered with, resulting in low security. Summary of the Invention

[0004] This application provides a blockchain-based copyright data protection method, system, device, and storage medium, aiming to improve the robustness of watermarks while enhancing the security of digital watermarks.

[0005] This application provides a method for protecting copyright data, the method comprising:

[0006] Multi-layer compression and sensing processing is performed on the original copyright data to obtain a feature watermark;

[0007] The target watermark is obtained by fusing the feature watermark and the real watermark, wherein the real watermark and the feature watermark are of the same size;

[0008] The target watermark is uploaded to the distributed storage system, and the hash value corresponding to the target watermark is uploaded to the blockchain.

[0009] Optionally, the step of performing multi-layer compressed sensing processing on the original copyright data to obtain the feature watermark includes:

[0010] Multi-layer compression and sensing processing is performed on the original copyright data to obtain visual privacy-protected copyright data;

[0011] The visual privacy protection copyright data is divided into multiple data blocks, and the characteristics of each data block are determined;

[0012] The feature watermark is generated based on the characteristics of each data block.

[0013] Optionally, the step of performing multi-layer compressed sensing processing on the original copyright data to obtain visual privacy-protected copyright data includes:

[0014] The original copyright data is divided into multiple data blocks, and the variance of each data block is calculated;

[0015] Construct a variance weight matrix based on the variance of each data block;

[0016] The adaptive measurement matrix is ​​obtained based on the variance weight matrix and the sparse binary matrix;

[0017] The weight value corresponding to each data block is determined based on the variance of each data block;

[0018] The visual privacy protection copyright data is obtained based on the weight value corresponding to each data block and the adaptive measurement matrix.

[0019] Optionally, the step of dividing the visual privacy protection copyright data into multiple data blocks and determining the characteristics of each data block includes:

[0020] The visual privacy protection copyright data is divided into multiple data blocks, and each data block is converted into a row vector;

[0021] Randomly generate column vectors using chaotic sequences;

[0022] The row vector corresponding to each data block is multiplied by the column vector to obtain the feature of each data block.

[0023] Optionally, the step of fusing the feature watermark and the real watermark to obtain the target watermark includes:

[0024] Extract the attribute information and operation information of the original copyright data;

[0025] Based on the attribute information and the operation information, a unique identifier for the original copyright data is generated;

[0026] The target watermark is obtained by fusing the unique identifier, the feature watermark, and the real watermark.

[0027] Optionally, after the steps of uploading the target watermark to the distributed storage system and uploading the hash value corresponding to the target watermark to the blockchain, the method further includes:

[0028] Determine the target hash value based on the contract address generated during the current transaction process;

[0029] Obtain the watermark corresponding to the target hash value from the distributed storage system;

[0030] The actual watermark is obtained based on the watermark corresponding to the target hash value and the feature watermark;

[0031] The actual watermark is compared with the real watermark to verify whether the watermark has been tampered with.

[0032] Optionally, after the steps of uploading the target watermark to the distributed storage system and uploading the hash value corresponding to the target watermark to the blockchain, the method further includes:

[0033] When a request to view copyright data is received, the copyright data to be viewed is determined based on the request.

[0034] Determine the importance level of the copyright data to be viewed;

[0035] Viewing permissions for the copyright data to be viewed are determined based on its importance level.

[0036] Furthermore, to achieve the above objectives, the present invention also provides a copyright data protection system comprising:

[0037] The multi-layer compressed sensing processing module is used to perform multi-layer compressed sensing processing on the original copyright data to obtain feature watermarks;

[0038] A fusion module is used to fuse the feature watermark and the real watermark to obtain the target watermark, wherein the real watermark and the feature watermark are of the same size;

[0039] The upload module is used to upload the target watermark to the distributed storage system and upload the hash value corresponding to the target watermark to the blockchain.

[0040] In addition, to achieve the above objectives, the present invention also provides a copyright data protection device comprising: a memory, a processor, and a copyright data protection program stored in the memory and executable on the processor, wherein the copyright data protection program, when executed by the processor, implements the steps of the copyright data protection method described above.

[0041] In addition, to achieve the above objectives, the present invention also provides a computer-readable storage medium having a copyright data protection program stored thereon, wherein the copyright data protection program, when executed by a processor, implements the steps of the copyright data protection method described above.

[0042] This application provides a blockchain-based copyright data protection method, system, device, and storage medium. Compared to related technologies that directly embed watermarks into the original copyright data, requiring modification of the original copyright data and resulting in data distortion and poor robustness, this application employs a multi-layer compression and sensing processing approach to first obtain a feature watermark from the original copyright data; then, it fuses the feature watermark with the real watermark to obtain the target watermark, thus improving the watermark's robustness. Furthermore, after obtaining the target watermark, it is uploaded to a distributed storage system, and the corresponding hash value is uploaded to the blockchain. This blockchain-based distributed storage of the watermark effectively avoids the risks of watermark attacks, theft, and tampering, improving the security of copyright data. Attached Figure Description

[0043] Figure 1 This is a flowchart illustrating the first embodiment of the copyright data protection method of the present invention;

[0044] Figure 2 A schematic diagram of the multi-layer compressed sensing processing flow for the adaptive block sensing weight matrix;

[0045] Figure 3 This is a functional block diagram of the copyright data protection system for this invention;

[0046] Figure 4 This is a schematic diagram of the structure of the device for protecting the copyright data of this invention.

[0047] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. The accompanying drawings are only one embodiment and not the entirety of the invention. Detailed Implementation

[0048] With increasing digitalization and informatization, copyrighted data, such as images, audio files, and video files, are being used more and more widely in social networks as carriers of information. During this widespread dissemination, there is a significant risk of information theft, making copyright data protection an increasingly prominent issue. Watermarking technology is a crucial technology for copyright data protection. Furthermore, blockchain technology has been recognized as a major force in various fields, enabling more secure and reliable digital watermarking. The popularity of blockchain technology stems primarily from its key characteristic as a secure, distributed underlying technology: the ability to add digital watermarks and accurate owner information without altering the original assets. Consequently, various digital watermarking algorithms have been proposed for copyright data protection.

[0049] However, with the development of technologies such as machine learning and artificial intelligence, copying other people's work has become increasingly easy. Existing technologies include generating hash values ​​for copyright data using permanent hash functions and storing them in the blockchain, and adding watermarks to the original copyright data using the traditional discrete cosine transform method. However, this approach is not only inefficient but also lacks robustness.

[0050] Therefore, neither digital watermarking nor blockchain can solve the aforementioned problems alone. Simply embedding the owner's identity into any digital asset using digital watermarking necessitates relying on authoritative third-party institutions to maintain records of asset ownership. This system becomes even more complex if the original owner decides to sell the asset. This over-reliance on third parties, which present single points of failure, means that every operation in the system consumes significant time, leading to inefficiency. Furthermore, simply using blockchain is not a good solution due to the high cost of storing information on it. Therefore, we need to combine the advantages of both digital watermarking and blockchain to create an efficient and stable system.

[0051] Previously, the original owner needed to provide copyright authorities with critical personal information, including copyright data or personally identifiable information. This information was then manually verified by a third-party organization and stored on a central server. This not only created a single point of failure but also made the entire system inefficient and costly due to the large amount of manual work involved. Furthermore, the risk of information tampering or leakage was also high.

[0052] Therefore, this invention plans to use blockchain to store copyright data. The immutable nature of blockchain ensures that once written, the record cannot be altered. The use of blockchain can ensure the effective maintenance of the entire lifecycle of assets, allowing any user to see the entire lifecycle of an asset at any point in time. Therefore, blockchain will help establish the authenticity of transactions, thereby proving the current owner of the asset.

[0053] The main technical solution of this application is as follows: Multi-layer compressed sensing processing is performed on the original copyright data to obtain a feature watermark; the feature watermark and the real watermark are fused to obtain a target watermark, wherein the real watermark and the feature watermark have the same size; the target watermark is uploaded to a distributed storage system, and the hash value corresponding to the target watermark is uploaded to a blockchain. The blockchain's storage structure can achieve secure storage of uploaded data, prevent information leakage and infringement, and prevent visual privacy-protected images from being tampered with. Furthermore, the technique of combining the real watermark with the feature watermark to obtain the target watermark effectively solves the problem of distortion of the original copyright data caused by direct watermark embedding, and improves the robustness of the watermark.

[0054] To better understand the above technical solutions, exemplary embodiments of this disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of this disclosure to those skilled in the art.

[0055] like Figure 1 As shown, in the first embodiment of this application, the method for protecting copyright data includes the following steps:

[0056] Step S110: Perform multi-layer compressed sensing processing on the original copyright data to obtain the feature watermark.

[0057] In this embodiment, the original copyright data of this application includes, but is not limited to, images, videos, audio, and files, as well as other copyrighted data. The feature watermark is a watermark possessing the features of the original copyright data. Multilayer compressed sensing processing is a signal sampling and reconstruction technique that combines the ideas of compressed sensing and multilayer decomposition. Compressed sensing is a technique that recovers the complete signal by sampling partial information of a signal and using a sparse representation. In traditional compressed sensing, the signal is usually represented as coefficients in a sparse domain and then sampled through a measurement matrix. However, for some types of signals, such as images and videos, their sparse representation may be difficult to meet sampling requirements. Therefore, multilayer compressed sensing processing introduces multi-level signal decomposition to increase the sparsity of the signal. Multilayer compressed sensing processing of the original copyright data includes the following steps:

[0058] (1) Multi-level decomposition: The input original copyright data is decomposed into multiple sub-bands (or frequency bands) through a series of transformations and filters. These sub-bands can be small blocks in the spatial domain, coefficients in the wavelet domain, etc.

[0059] (2) Compressed sensing sampling: In each sub-band, the original copyright data is sampled using a measurement matrix to obtain a small number of measurement values.

[0060] (3) Sparse representation and reconstruction: For the measurement values ​​of each subband, a sparse representation method, such as a dictionary-based method or an optimization algorithm, is used to recover the sparse representation of the original copyright data.

[0061] (4) Reconstruction and restoration: The sparse representations in each subband are merged and inversely transformed to obtain the reconstructed complete copyright data.

[0062] The advantage of multilayer compressed sensing processing lies in its ability to utilize the sparsity of signals across multiple scales, thereby improving the quality and efficiency of reconstruction. It can be applied to various fields, such as image processing, video compression, and medical imaging, to reduce sampling rates and storage requirements while maintaining good reconstruction quality. This application employs multilayer compressed sensing processing technology to process the original copyright data, effectively reducing redundancy in the data processing process and improving data processing efficiency and quality.

[0063] Step S120: Fuse the feature watermark and the real watermark to obtain the target watermark, wherein the real watermark and the feature watermark have the same size.

[0064] In this embodiment, a feature watermark is an invisible piece of information embedded in digital media (such as images, audio, and video) for purposes such as identity verification, copyright protection, and data tracking. Feature watermarks are characterized by invisibility, robustness, and security. The real watermark is designed by the copyright owner, and its size is identical to the real watermark, allowing them to overlap.

[0065] In this embodiment, after obtaining the feature watermark, the feature watermark and the real watermark are fused to obtain the target watermark, so that the target watermark embeds the aforementioned invisible information, thereby achieving the protection of copyright data.

[0066] Step S130: Upload the target watermark to the distributed storage system and upload the hash value corresponding to the target watermark to the blockchain.

[0067] In this embodiment, after obtaining the target watermark, it is uploaded to a distributed storage system. After uploading the target watermark to the distributed storage system, the hash value corresponding to the target watermark is returned. Uploading the returned hash value to the blockchain enables tamper-proofing of the watermark. Specifically, through a smart contract, the target watermark is uploaded to the blockchain, obtaining a hash value. The hash value is the contract address of the transaction. Based on this contract address, the corresponding watermark can be viewed.

[0068] In this embodiment, the hash value can be generated not only based on the target watermark, but also by combining the user's email address and public key.

[0069] In other embodiments, a link may be provided to copyright users to download the encrypted target watermark.

[0070] In one application scenario, the system includes copyright owners and copyright verifiers. The copyright owner is the creator of the original copyright data, who first registers the asset on the blockchain; verifiers are any members of the public who want to verify the originality of the copyright data. After the copyright data is created, the copyright owner can transfer it to other users, and the corresponding transaction records will be recorded on the blockchain.

[0071] If the original copyright owner wishes to transfer ownership of copyright data, the transaction will be conducted on the blockchain. During the asset transfer, the original copyright data is first retrieved from the distributed storage system, IPFS, and then the watermark is extracted. The original watermark containing the previous original copyright owner's information is replaced by a new watermark containing the new copyright owner's information. Similar changes are made within IPFS. Simultaneously, a new transaction is added to the blockchain to reflect this change in asset ownership, ensuring that the blockchain always contains information about the asset's latest owner. This information is permanently stored on the blockchain, creating an auditable log that provides access to the asset's complete lifecycle, traceable at any time by anyone. Furthermore, smart contracts ensure that all information is automatically written to the blockchain, eliminating reliance on a central authority and effectively reducing the overhead of maintaining and physically verifying records.

[0072] Compared to current systems involving third parties, this application operates much faster, and thanks to blockchain technology, trust between buyers and sellers involved in the assets is automatically established without the need for any external parties. However, since all information is public, data manipulation is highly possible. Storing watermark information on the blockchain ensures that assets cannot be copied, and any changes to the watermark information can be easily identified. This method can also be extended to include multiple watermarks, and the timestamping feature of the blockchain helps track the order in which each watermark was added. Storing all assets such as audio files, video files, or images directly on the blockchain is not very efficient because each transaction and storage costs money. Therefore, decentralized storage services, such as IPFS, can be used to store these assets. Decentralized storage ensures that no single entity has complete control over the assets, so no single entity can manipulate the asset's information. Simultaneously, storing the asset's address on the blockchain ensures that the asset's address remains permanently stored on the network.

[0073] This embodiment, based on the above technical solution, improves the robustness of the watermark by first performing multi-layer compression and sensing processing on the original copyright data to obtain a feature watermark; then fusing the feature watermark and the real watermark. Furthermore, after obtaining the target watermark, it is uploaded to a distributed storage system, and the hash value corresponding to the target watermark is uploaded to the blockchain. This blockchain-based distributed storage of the watermark effectively avoids the risks of attacks, theft, and tampering with the watermark.

[0074] Further, based on step S110 of the first embodiment, performing multi-layer compressed sensing processing on the original copyright data to obtain the feature watermark includes: performing multi-layer compressed sensing processing on the original copyright data to obtain visual privacy-protected copyright data; and then generating the feature watermark based on the visual privacy-protected copyright data. Here, the visual privacy-protected copyright data refers to copyright data in a visual privacy-protected state, that is, copyright data generated using visual privacy protection technology. Since the original copyright data may contain privacy content such as personal identity, location, and behavior, this privacy content is crucial for protecting individual rights and social security. Therefore, a series of processing steps are needed to obtain visual privacy-protected copyright data to protect privacy. Performing multi-layer compressed sensing processing on the original copyright data to obtain visual privacy-protected copyright data can protect personal privacy and sensitive information.

[0075] Optionally, performing multi-layer compressed sensing processing on the original copyright data to obtain the feature watermark specifically includes the following steps:

[0076] Step S111: Perform multi-layer compressed sensing processing on the original copyright data to obtain visual privacy-protected copyright data.

[0077] In this embodiment, combining multi-layer compressed sensing and visual privacy protection technology can further enhance the protection of personal privacy.

[0078] Optionally, refer to Figure 2 The process of performing multi-layer compression and perception processing on the original copyright data to obtain visual privacy-protected copyright data includes the following steps:

[0079] Step S1111: Divide the original copyright data into multiple data blocks and calculate the variance of each data block.

[0080] In this embodiment, taking the original copyright data as image data and data blocks as image blocks as an example, the image is divided into blocks, and the variance value of each image block is calculated. The adaptive block-aware weight matrix is ​​an improved matrix obtained by supplementing the adaptive measurement matrix with image block variance information. Its idea is to extract more information from blocks with smaller variances, and to reduce the extraction of information from blocks with larger variances. The formula for calculating the variance of each data block is as follows:

[0081] .

[0082] in, The mean; For the first in the image Line number The coefficients corresponding to the columns are obtained through image patch transformation, with the image patch size being [value missing]. .

[0083] Step S1112: Construct a variance weight matrix based on the variance of each data block.

[0084] In this embodiment, the variance of each data block can be obtained using formula (1). The variance of each data block is then placed on the first diagonal of the matrix to obtain the variance weight matrix:

[0085] .

[0086] in, and These represent the number of rows and columns in an image patch, respectively.

[0087] Step S1113: Obtain the adaptive measurement matrix based on the variance weight matrix and the sparse binary matrix.

[0088] In this embodiment, the adaptive measurement matrix can be obtained by multiplying the variance weight matrix and the sparse binary matrix. Specifically, the compressed sampling formula can be derived from the general compressed sensing formula. The compression sampling formula is as follows: Figure 2 The adaptive measurement matrix is ​​obtained by multiplying the sparse binary moments (sparse binary matrix) and the weight moments (variance weight matrix). In this compressed sampling formula, S is the image patch information obtained after vectorization. This is the vector obtained after processing by discrete cosine transform; This represents a sparse binary matrix. Compressed sampling is introduced into the image data to obtain a visually privacy-preserving image, thus protecting image privacy from the source.

[0089] Step S1114: Determine the weight value corresponding to each data block based on the variance of each data block.

[0090] In this embodiment, the weight value corresponding to each data block is denoted as . The result is obtained by formula (3):

[0091]

[0092] In formula (3), the denominator is the sum of the variances of all data blocks, and the numerator is the variance of the Kth data block.

[0093] Step S1115: Obtain the visual privacy protection copyright data based on the weight value corresponding to each data block and the adaptive measurement matrix.

[0094] In this embodiment, an adaptive weight matrix is ​​obtained based on the weight value corresponding to each data block and the adaptive measurement matrix. Then, the visual privacy protection copyright data is obtained based on the adaptive weight matrix. The adaptive weight matrix is ​​an improved matrix obtained by supplementing the adaptive measurement matrix with the variance information of the image blocks, denoted as... , as in formula (4).

[0095]

[0096] Where L is the total number of blocks in the image. M represents the total number of samples in each image patch. This represents the initial number of image patches measured.

[0097] This embodiment addresses the issue that, in related technologies, adding watermarks to original copyright data using the discrete cosine transform method requires altering the original copyright data, leading to reduced efficiency and poor robustness of the watermark. This application proposes a multilayer compressed sensing technique based on an adaptive block-aware weight matrix. This technique uses matrix calculations to perform sparse sampling on the image to obtain feature watermarks, effectively reducing redundancy and improving processing efficiency. The adaptive block-aware weight matrix is ​​a technique used for compressed sensing image reconstruction. Traditional compressed sensing typically uses sparse transform domain representations, such as wavelet transform or sparse dictionary representations, to obtain a sparse representation of the signal. The signal is then reconstructed by sparse optimization of the observed data. The adaptive block-aware weight matrix introduces the concept of an adaptive weight matrix to improve reconstruction quality.

[0098] Alternatively, the original copyright data can be subjected to multi-layer compressed sensing processing using any of the following methods to obtain visual privacy-protected copyright data:

[0099] (1) Region-selective perception: In the process of multi-layer compressed perception, different regions can be selectively perceived and reconstructed based on the features of copyright data, such as images or videos. By defining sensitive regions in advance and only perceiving and reconstructing non-sensitive regions, sensitive information can be avoided from being directly obtained.

[0100] (2) Region occlusion and blurring: In terms of visual privacy protection, region occlusion or blurring techniques can be used to hide sensitive areas. In multilayer compressed sensing, unselected sensitive areas can be occluded or blurred to reduce the leakage of sensitive information.

[0101] (3) Reduce sampling rate: A core concept in multilayer compressed sensing is sparse representation, which reduces the sampling rate by using a sparse representation of the measurement signal. When combined with visual privacy protection technology, the sampling rate can be further reduced according to privacy requirements to reduce the exposure of raw data.

[0102] (4) Encryption and secure transmission: Encryption and secure transmission are important technical means for protecting visual privacy. Encryption algorithms can be introduced into the multi-layer compressed sensing process to ensure the security of image or video data during transmission and storage, and to prevent unauthorized access.

[0103] (5) Access control and permission management: To further protect personal privacy, an effective permission management mechanism can be established by combining multi-layer compression sensing and access control technologies. Only authorized personnel can access and reconstruct image or video data, thereby ensuring data privacy.

[0104] After performing compressed sensing processing on the original copyright data to obtain compressed and encrypted visual privacy-protected copyright data, step S112 is executed:

[0105] Step S112: Divide the visual privacy protection copyright data into multiple data blocks and determine the characteristics of each data block.

[0106] In this embodiment, the visual privacy-protected copyright data can be a visual privacy-protected image, a visual privacy-protected video, or something else. The data block can be a corresponding image block, video block, or something else. If the visual privacy-protected copyright data is a visual privacy-protected image, the visual privacy-protected image is divided into multiple image blocks, and the characteristics of each image block are determined. If the visual privacy-protected copyright data is a visual privacy-protected video, the visual privacy-protected video is divided into multiple video blocks, and the characteristics of each video block are determined.

[0107] Optionally, dividing the visual privacy protection copyright data into multiple data blocks and determining the characteristics of each data block includes the following steps:

[0108] Step S1121: Divide the visual privacy protection copyright data into multiple data blocks and convert each data block into a row vector;

[0109] In this embodiment, all the data blocks are the same size and do not overlap. For example, the visual privacy-preserving image is divided into 8×8 non-overlapping image blocks. Each image block is converted into a 1×64 row vector.

[0110] Step S1122: Randomly generate column vectors using chaotic sequences.

[0111] In this embodiment, a 64×1 column vector is randomly generated using a chaotic sequence.

[0112] Step S1123: Multiply the row vector corresponding to each data block with the column vector to obtain the feature of each data block.

[0113] In this embodiment, the row vector and column vector corresponding to each image patch are multiplied to obtain the features of each image patch, and thus the feature value of each image patch can be obtained. Specifically, the 1×64 row vector and the 64... Multiplying column vectors of 1 together yields a 1. A vector of 1 represents the feature value corresponding to the image patch.

[0114] Step S113: Generate the feature watermark based on the characteristics of each data block.

[0115] In this embodiment, the feature watermark is directly obtained from the feature matrix formed by the feature values ​​of each image block. The aforementioned feature watermark is an invisible information embedded in digital media (such as images, audio, and video) for purposes such as identity verification, copyright protection, and data tracking. Feature watermarks possess characteristics such as invisibility, robustness, and security.

[0116] Furthermore, based on the above embodiments, in order to ensure the uniqueness of the target watermark, the process of fusing the feature watermark and the real watermark to obtain the target watermark includes the following steps:

[0117] Step S121: Extract the attribute information and operation information of the original copyright data.

[0118] In this embodiment, when the original copyright data is an image, the attribute information of the original copyright data includes image pixel information, image resolution, image size, etc., and the operation information includes operator information, the organization or individual to which the image belongs, etc. When the original copyright data is a video, the attribute information of the original copyright data includes video duration, video refresh rate, video quality, etc., and the operation information includes historical viewing count, the organization or individual to which the video belongs, etc.

[0119] Step S122: Generate a unique identifier for the original copyright data based on the attribute information and the operation information.

[0120] In this embodiment, if the data types of the attribute information and the operation information are inconsistent, the attribute information and the operation information are converted into data with the same data type. The converted attribute information and the operation information are then concatenated to obtain a unique identifier for the original copyright data. This unique identifier can then be used to locate relevant information about the original copyright data.

[0121] Step S123: Fuse the unique identifier, the feature watermark, and the real watermark to obtain the target watermark.

[0122] In this embodiment, after obtaining the unique identifier, the unique identifier and the real watermark are embedded into the feature watermark as a digital watermark to obtain the target watermark. Because the unique identifier is embedded in the feature watermark, the uniqueness of the target watermark is guaranteed. The unique identifier is also stored on the blockchain.

[0123] Furthermore, based on the above embodiments, after the steps of uploading the target watermark to the distributed storage system and uploading the hash value corresponding to the target watermark to the blockchain, the following steps are also included:

[0124] Step S210: Determine the target hash value based on the contract address generated during the current transaction process;

[0125] Step S220: Obtain the watermark corresponding to the target hash value from the distributed storage system;

[0126] Step S230: Obtain the actual watermark based on the watermark corresponding to the target hash value and the feature watermark;

[0127] Step S240: Compare the actual watermark with the real watermark to verify whether the watermark has been tampered with.

[0128] In this embodiment, a file is uploaded to the blockchain via a smart contract, resulting in a hash value. The hash value is the contract address of the transaction. When extracting the watermark file from the blockchain, the corresponding smart contract is located by inputting the contract address, and the target watermark is downloaded from the distributed storage network using the hash value of the watermark file stored in the contract. This target watermark is then combined with the feature watermark calculated from the compressed and encrypted image to obtain the actual watermark. The integrity and validity of the compressed sensing image are verified by comparing the extracted actual watermark with the real watermark from step S120.

[0129] Furthermore, based on the above embodiments, after the steps of uploading the target watermark to the distributed storage system and uploading the hash value corresponding to the target watermark to the blockchain, the method further includes:

[0130] Step S310: When receiving a request to view copyright data, determine the copyright data to be viewed based on the request;

[0131] Step S320: Determine the importance level of the copyright data to be viewed;

[0132] Step S330: Determine the viewing permissions for the copyright data to be viewed based on its importance level.

[0133] In this embodiment, to better coordinate the blockchain watermark with actual conditions and to improve copyright data protection, when a user views copyright data, it is necessary to assess the importance level of the copyright data the user intends to view to verify whether the user has the necessary permissions, thus protecting the copyright data. Specifically, a model for assessing the importance of the watermark can be created, and the importance level of the copyright data the user intends to view can be determined based on this model. The construction process of the copyright data importance assessment model is as follows:

[0134] (1) Feature selection. The webpage probe captures relevant features reflecting the importance of copyright data, such as the number of users requesting the copyright data, the number of transfers, and the number of collections and likes, and then normalizes these features.

[0135] (2) Data collection. Collect copyrighted data that has been watermarked;

[0136] (3) Model training. Based on the selected features and collected copyright data, an image importance assessment model is established. The convolutional neural network method in machine learning is used to train the model, which is trained on training data. During the training process, the parameters of the model are continuously adjusted so that the model can better fit the data.

[0137] (4) Model evaluation. The model is evaluated using test set data, and indicators such as prediction accuracy and F1 score are calculated.

[0138] (5) Model optimization. Based on the model evaluation results, the model is continuously optimized by adding features and adjusting parameters to improve prediction accuracy;

[0139] (6) Model Application. The trained copyright data importance assessment model is applied to specific scenarios. By analyzing the existing copyright data on web pages, the importance level of the copyright data is classified according to the scores of the copyright data output by the network.

[0140] In this embodiment, the importance level of the copyright data to be viewed can be divided into three levels: high (score between 86-100), medium (score between 60-85), and low (score between 0-84). The viewing permissions and verification method for the copyright data to be viewed can be determined according to its importance level, as follows:

[0141] (1) For high-level copyright data, users need to perform the following operations: ① Real-name authentication; ② After the authentication is passed, the relevant introduction of the copyright data can be viewed; ③ If you need to view the copyright data, you need to apply for authorization from the copyright owner and sign the relevant copyright protection terms. You can only view it after the review is approved.

[0142] (2) For medium-level copyright data, users need to perform the following operations: ① Real-name authentication; ② After successful authentication, users can view the copyright data and related information. If it is necessary to obtain copyright data, it is necessary to apply for authorization from the copyright owner.

[0143] (3) For low-level copyright data, users need to perform the following operations: ① They can view the copyright data and related introductions without real-name authentication; ② If they need to obtain the copyright data, they need to perform real-name authentication first, and apply for authorization from the copyright owner after the authentication is approved.

[0144] Therefore, different viewing permissions are granted based on the importance level of different copyrighted data, which can meet the usage needs of different application scenarios.

[0145] Furthermore, to ensure superior security for the application, a webpage probe is added to monitor the application pages for data collection commands such as screenshots, preventing users from obtaining image data through unauthorized means. This webpage probe primarily has the following functionalities:

[0146] (1) By capturing network data packets of web page requests and related responses, analyze the relevant information such as HTTP requests, response headers, request bodies and response bodies contained in the data packets; at the same time, search for keywords related to screenshots in the network data packets. The purpose of doing this is to determine whether there are information collection instructions such as screenshot instructions in the current web page behavior. Once behaviors containing these keywords are found, they are judged to be illegal behaviors.

[0147] (2) Monitoring browser extensions: In practical applications, users often use third-party extensions to complete information collection operations such as screenshots. Therefore, this webpage probe can monitor the installation, use and usage of browser extensions. Once detected, it is judged as a violation.

[0148] (3) Regularly check webpage code: Set a time threshold to enable the webpage probe to regularly check the webpage source code to find whether there is code for collecting information such as screenshots in the current code.

[0149] Using the above functions of the webpage probe, once a user is found to be engaging in illegal behavior, a pop-up warning will be displayed. If the user still insists on performing the relevant operations, the page will be forcibly closed to protect data copyright.

[0150] This invention provides an embodiment of a method for protecting copyright data. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown here.

[0151] like Figure 3 As shown, the copyright data protection system provided in this application includes:

[0152] The multi-layer compression sensing processing module 10 is used to perform multi-layer compression sensing processing on the original copyright data to obtain feature watermarks;

[0153] The fusion module 20 is used to fuse the feature watermark and the real watermark to obtain the target watermark, wherein the real watermark and the feature watermark are of the same size;

[0154] The upload module 30 is used to upload the target watermark to the distributed storage system and upload the hash value corresponding to the target watermark to the blockchain.

[0155] The specific implementation of the copyright data protection system of this invention is basically the same as the embodiments of the copyright data protection method described above, and will not be repeated here.

[0156] like Figure 4 As shown, Figure 4 This is a schematic diagram of the hardware operating environment of the copyright data protection device according to an embodiment of the present invention. The copyright data protection device may include: a processor 1001, such as a CPU; a memory 1005; a user interface 1003; a network interface 1004; and a communication bus 1002. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen and an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed RAM memory or a stable memory, such as a disk storage device. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0157] Those skilled in the art will understand that Figure 4 The copyright data protection device structure shown does not constitute a limitation on the copyright data protection device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0158] like Figure 4 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a network communication module, a user interface module, and a copyright data protection program. The operating system is a program that manages and controls the hardware and software resources of the copyright data protection device, as well as the execution of the copyright data protection program and other software or programs.

[0159] exist Figure 4 In the copyright data protection device shown, the user interface 1003 is mainly used to connect to the terminal and communicate with the terminal; the network interface 1004 is mainly used to communicate with the backend server; and the processor 1001 can be used to call the copyright data protection program stored in the memory 1005.

[0160] In this embodiment, the copyright data protection device includes: a memory 1005, a processor 1001, and a copyright data protection program stored in the memory and executable on the processor, wherein:

[0161] When processor 1001 calls the copyright data protection program stored in memory 1005, it performs the following operations:

[0162] Multi-layer compression and sensing processing is performed on the original copyright data to obtain a feature watermark;

[0163] The target watermark is obtained by fusing the feature watermark and the real watermark, wherein the real watermark and the feature watermark are of the same size;

[0164] The target watermark is uploaded to the distributed storage system, and the hash value corresponding to the target watermark is uploaded to the blockchain.

[0165] Based on the same inventive concept, this application also provides a computer-readable storage medium storing a copyright data protection program. When the copyright data protection program is executed by a processor, it implements the various steps of the copyright data protection method described above and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0166] Since the storage medium provided in this application embodiment is the storage medium used to implement the method of this application embodiment, those skilled in the art can understand the specific structure and variations of the storage medium based on the method described in this application embodiment, and therefore will not be repeated here. All storage media used in the method of this application embodiment are within the scope of protection of this application.

[0167] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0168] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0169] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, television, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0170] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for protecting copyright data, characterized in that, The methods for protecting the copyright data include: Multi-layer compression and sensing processing is performed on the original copyright data to obtain a feature watermark; The target watermark is obtained by fusing the feature watermark and the real watermark, wherein the real watermark and the feature watermark are of the same size, and the real watermark is designed by the copyright owner; The target watermark is uploaded to the distributed storage system, and the hash value corresponding to the target watermark is uploaded to the blockchain; The process of fusing the feature watermark and the real watermark to obtain the target watermark includes: Extract the attribute information and operation information of the original copyright data; Based on the attribute information and the operation information, a unique identifier for the original copyright data is generated; The target watermark is obtained by fusing the unique identifier, the feature watermark, and the real watermark.

2. The method for protecting copyright data as described in claim 1, characterized in that, The step of performing multi-layer compression sensing processing on the original copyright data to obtain the feature watermark includes: Multi-layer compression and sensing processing is performed on the original copyright data to obtain visual privacy-protected copyright data; The visual privacy protection copyright data is divided into multiple data blocks, and the characteristics of each data block are determined; The feature watermark is generated based on the characteristics of each data block.

3. The method for protecting copyright data as described in claim 2, characterized in that, The step of performing multi-layer compressed sensing processing on the original copyright data to obtain visual privacy-protected copyright data includes: The original copyright data is divided into multiple data blocks, and the variance of each data block is calculated; Construct a variance weight matrix based on the variance of each data block; The adaptive measurement matrix is ​​obtained based on the variance weight matrix and the sparse binary matrix; The weight value corresponding to each data block is determined based on the variance of each data block; The visual privacy protection copyright data is obtained based on the weight value corresponding to each data block and the adaptive measurement matrix.

4. The method for protecting copyright data as described in claim 2, characterized in that, The step of dividing the visual privacy protection copyright data into multiple data blocks and determining the characteristics of each data block includes: The visual privacy protection copyright data is divided into multiple data blocks, and each data block is converted into a row vector; Randomly generate column vectors using chaotic sequences; The row vector corresponding to each data block is multiplied by the column vector to obtain the feature of each data block.

5. The method for protecting copyright data as described in claim 1, characterized in that, After the steps of uploading the target watermark to the distributed storage system and uploading the hash value corresponding to the target watermark to the blockchain, the method further includes: Determine the target hash value based on the contract address generated during the current transaction process; Obtain the watermark corresponding to the target hash value from the distributed storage system; The actual watermark is obtained based on the watermark corresponding to the target hash value and the feature watermark; The actual watermark is compared with the real watermark to verify whether the watermark has been tampered with.

6. The method for protecting copyright data as described in claim 1, characterized in that, After the steps of uploading the target watermark to the distributed storage system and uploading the hash value corresponding to the target watermark to the blockchain, the method further includes: When a request to view copyright data is received, the copyright data to be viewed is determined based on the request. Determine the importance level of the copyright data to be viewed; Viewing permissions for the copyright data to be viewed are determined based on its importance level.

7. A copyright data protection system, characterized in that, The copyright data protection system includes: The multi-layer compressed sensing processing module is used to perform multi-layer compressed sensing processing on the original copyright data to obtain feature watermarks; A fusion module is used to fuse the feature watermark and the real watermark to obtain the target watermark, wherein the real watermark and the feature watermark are of the same size, and the real watermark is designed by the copyright owner; The upload module is used to upload the target watermark to the distributed storage system and upload the hash value corresponding to the target watermark to the blockchain; The process of fusing the feature watermark and the real watermark to obtain the target watermark includes: Extract the attribute information and operation information of the original copyright data; Based on the attribute information and the operation information, a unique identifier for the original copyright data is generated; The target watermark is obtained by fusing the unique identifier, the feature watermark, and the real watermark.

8. A copyright data protection device, characterized in that, The copyright data protection device includes: a memory, a processor, and a copyright data protection program stored in the memory and running on the processor, wherein when the copyright data protection program is executed by the processor, it implements the steps of the copyright data protection method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, It stores a copyright data protection program, which, when executed by a processor, implements the steps of the copyright data protection method according to any one of claims 1-6.

Images