Intelligent manufacturing data security protection method based on post-quantum technology

Abstract

This invention belongs to the field of data security protection technology, specifically relating to a data security protection method for intelligent manufacturing based on post-quantum technology. The method includes: dividing a production monitoring image into multiple non-overlapping blocks of a specified size using a block truncation algorithm based on absolute distance, and obtaining the bitmap, low-quantization value, and high-quantization value of each non-overlapping block; the specified size being a pre-selected size with the highest overall effect value; embedding secret information into the bitmap of each non-overlapping block using a block-based hiding algorithm according to a preset key matrix and weight matrix, obtaining the ciphertext bitmap of each non-overlapping block; encoding the low-quantization value, high-quantization value, and ciphertext bitmap of each non-overlapping block, and encrypting and storing the encoding result based on the NIST quantum-resistant cryptography standard algorithm. This invention not only achieves efficient integration of compression, encryption, and watermark embedding, but also maximizes overall performance while ensuring image quality.

Description

Technical Field

[0001] This invention relates to the field of data security protection technology. More specifically, this invention relates to a data security protection method for intelligent manufacturing based on post-quantum technology. Background Technology

[0002] In the process of intelligent manufacturing, production monitoring images are acquired in real time through various sensors, cameras and other data acquisition devices to monitor the operating status of production equipment, product quality and changes in the production environment.

[0003] With the rapid development of intelligent manufacturing, production monitoring images are playing an increasingly important role in industrial processes. They are not only an important foundation for realizing intelligent production and decision-making, but also provide key information for fault diagnosis, quality control and optimized scheduling.

[0004] However, since production monitoring images often contain a large amount of sensitive data, ensuring their security during transmission and storage has become an urgent problem to be solved.

[0005] Meanwhile, considering the large amount of data in production monitoring images, direct storage or transmission would lead to excessive resource consumption, so compression processing is necessary.

[0006] In addition, to prevent images from being tampered with or forged, secret information needs to be embedded to verify their integrity and authenticity.

[0007] While traditional encryption, compression, and information hiding technologies perform well in their respective fields, they pose security risks when facing quantum computing threats. Furthermore, existing information hiding technologies are often independent of compression and encryption processes, making it difficult to simultaneously meet the requirements of efficient compression, security protection, and tamper-proofing, and thus failing to achieve efficient integrated processing. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention provides a smart manufacturing data security protection method based on post-quantum technology, comprising: dividing a production monitoring image into multiple non-overlapping blocks of a specified size using a block truncation algorithm based on absolute distance, and obtaining the bitmap, low quantization value, and high quantization value of each non-overlapping block; embedding secret information into the bitmap of each non-overlapping block using a block hiding algorithm according to a preset key matrix and weight matrix; encoding the obtained ciphertext bitmap, low quantization value, and high quantization value of each non-overlapping block; encrypting and storing the encoding result based on the NIST quantum-resistant cryptography standard algorithm; wherein, the specified size is a pre-selected size with the largest comprehensive effect value, and the comprehensive effect value of each pre-selected size... The method for obtaining the image compression degree corresponding to the pre-selected size is as follows: Calculate the image compression degree corresponding to the pre-selected size based on the number of non-overlapping blocks corresponding to the pre-selected size and the amount of data stored for each non-overlapping block; calculate the loss caused by compression coding corresponding to the pre-selected size based on the difference between the grayscale value of the pixels in all non-overlapping blocks corresponding to the pre-selected size and the low quantization value and the high quantization value; calculate the loss caused by information hiding corresponding to the pre-selected size based on the number of times the secret information is embedded in all non-overlapping blocks corresponding to the pre-selected size and the difference between the low quantization value and the high quantization value; and weight the image compression degree, the loss caused by information hiding, and the loss caused by compression coding corresponding to each pre-selected size to obtain the comprehensive effect value of each pre-selected size.

[0009] This invention employs a block truncation algorithm based on absolute distance to perform lossy compression on production monitoring images, effectively reducing the data volume. Simultaneously, a block hiding algorithm embeds secret information into the bitmaps of each non-overlapping block, ensuring the authenticity and integrity of the image. Furthermore, the NIST quantum-resistant cryptography standard algorithm is used to encrypt and store low-quantization values, high-quantization values, and the ciphertext bitmaps after information hiding, greatly enhancing the security of the image data and enabling it to maintain its anti-attack capability even in a quantum computing environment. This invention not only achieves seamless integration of compression, encryption, and information hiding but also considers compression efficiency, security protection, and anti-tampering requirements, significantly reducing storage and transmission costs while ensuring the reliability and availability of production monitoring images in intelligent manufacturing processes.

[0010] Furthermore, by determining the specified size that maximizes the overall effect value, this invention can find the optimal balance between compression efficiency and image quality. It not only achieves efficient integration of compression, encryption, and watermark embedding, but also maximizes overall performance while ensuring image quality, providing a reliable solution for production monitoring image processing in intelligent manufacturing processes.

[0011] Preferably, obtaining the bitmap, low quantization value, and high quantization value of each non-overlapping block includes: calculating the average grayscale value of all pixels in the non-overlapping block; classifying pixels with grayscale values ​​greater than the average value into a first-class pixel and writing "1" at the corresponding position in the bitmap; classifying pixels with grayscale values ​​not greater than the average value into a second-class pixel and writing "0" at the corresponding position in the bitmap; thereby obtaining the bitmap of the non-overlapping block; calculating the average grayscale value of all first-class pixels and using the rounded result as the high quantization value of the non-overlapping block; calculating the average grayscale value of all second-class pixels and using the rounded result as the low quantization value of the non-overlapping block.

[0012] Preferably, the preset key matrix is ​​a matrix of size 4×4, and the elements in the matrix are 0 or 1.

[0013] Preferably, the weight matrix is ​​a 4×4 matrix, and the elements in the matrix are integers in the range [1,7], and each integer in the range [1,7] appears 2 or 3 times in the matrix.

[0014] Preferably, the step of embedding the secret information into the bitmap of each non-overlapping block using the block hiding algorithm includes: dividing the bitmap of the non-overlapping block into multiple sub-blocks of size 4×4; for any sub-block... , pair of blocks Perform the transformation to obtain the transformed sub-block. The transformed sub-blocks are required. satisfy In the formula, , These are the key matrix and the weight matrix, respectively. This indicates summation. This represents the decimal number corresponding to the secret information. This represents the XOR operation. This indicates digit-wise multiplication. Represents the modulo operation; the pair of sub-blocks Transformation refers to: transforming sub-blocks At least one element is swapped between 0 and 1.

[0015] Preferably, the preselected size is equal to ,in, , All are integers in the range [1, 5].

[0016] Preferably, the calculation of the image compression degree corresponding to the pre-selected size includes: In the formula, For pre-selected size, Indicates the pre-selected size The corresponding level of image compression, Indicates the pre-selected size The corresponding number of non-overlapping blocks, the amount of data stored in each non-overlapping block is equal to , To determine the size of the generated surveillance images.

[0017] In this invention, the calculation of the overall effect value includes the image compression degree as an indicator. Selecting the specified size with the largest overall effect value can ensure that the highest possible compression ratio is achieved while meeting certain image quality requirements, thereby effectively reducing storage and transmission costs.

[0018] Preferably, the calculation of the loss caused by compression encoding corresponding to the pre-selected size includes: In the formula, For pre-selected size, Indicates the pre-selected size The degree of loss caused by the corresponding compression encoding. Indicates the pre-selected size The corresponding number of non-overlapping blocks, , The first The number of all Class I pixels and all Class II pixels in a non-overlapping block , The first The first non-overlapping block The first type of pixel and the first The grayscale values ​​of two types of pixels , The first High and low quantization values ​​of non-overlapping blocks To produce the size of the surveillance images, This represents the function that takes the absolute value.

[0019] This invention analyzes the degree of loss caused by compression coding. The comprehensive effect value can fully reflect the impact of different block partitioning sizes on image quality. Selecting the specified size with the largest comprehensive effect value can minimize image distortion and visual quality degradation while ensuring compression efficiency.

[0020] Preferably, the calculation of the degree of loss caused by information hiding corresponding to the pre-selected size includes: In the formula, For pre-selected size, Indicates the pre-selected size The extent of loss caused by the corresponding information hiding, Indicates the pre-selected size The number of corresponding non-overlapping blocks, and the number of times the secret information of each non-overlapping block corresponding to the pre-selected size is embedded equal to... , , Pre-selected size The corresponding number High and low quantization values ​​of non-overlapping blocks To determine the size of the generated surveillance images.

[0021] In this invention, the number of times the secret information is embedded and the difference between low and high quantization values ​​directly affect the performance of information hiding. By optimizing the block partitioning size by calculating the degree of loss caused by information hiding, the impact of information hiding on image quality can be reduced, while ensuring the robustness and invisibility of watermark information, thereby better protecting the authenticity and integrity of image data.

[0022] Preferably, the step of weighting the image compression degree, the loss due to information hiding, and the loss due to compression coding for each pre-selected size to obtain a comprehensive effect value for each pre-selected size includes: In the formula, Indicates the pre-selected size The overall effect value, , , These represent the pre-selected dimensions. The corresponding image compression level, the degree of loss caused by compression coding, and the degree of loss caused by information hiding. Indicates the loss threshold. This represents the natural exponential function. , These are the first weight and the second weight, respectively. , All are greater than 0, and .

[0023] This invention introduces a weighted mechanism in the calculation of the overall effect value, which can adjust the weight parameters according to the specific application scenario and flexibly balance the relationship between compression efficiency, image quality and information hiding performance to meet the diverse needs in the intelligent manufacturing process.

[0024] The beneficial effects of the present invention are:

[0025] This invention not only achieves seamless integration of compression, encryption, and information hiding, but also takes into account the requirements of compression efficiency, security protection, and anti-tampering, significantly reducing storage and transmission costs while ensuring the reliability and availability of production monitoring images in the intelligent manufacturing process. Furthermore, by determining the specified size that maximizes the overall effect value, this invention can find the optimal balance between compression efficiency and image quality. It not only achieves efficient integration of compression, encryption, and watermark embedding, but also maximizes overall performance while ensuring image quality, providing a reliable solution for production monitoring image processing in the intelligent manufacturing process. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating the intelligent manufacturing data security protection method based on post-quantum technology in this invention;

[0027] Figure 2 This is a flowchart illustrating step S2;

[0028] Figure 3 This is a flowchart illustrating the method for obtaining the comprehensive effect value of each pre-selected dimension in step S21;

[0029] Figure 4 This is a schematic diagram illustrating non-overlapping blocks;

[0030] Figure 5 This is a schematic diagram illustrating a bitmap of non-overlapping blocks;

[0031] Figure 6 This is a schematic diagram illustrating a preset key matrix;

[0032] Figure 7 This is a schematic diagram illustrating the preset weight matrix;

[0033] Figure 8 This is a schematic diagram illustrating the transformed sub-blocks. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0036] This invention discloses a data security protection method for intelligent manufacturing based on post-quantum technology, referring to... Figure 1 This includes steps S1 to S2:

[0037] S1. Collect production monitoring images in intelligent manufacturing.

[0038] It should be noted that intelligent manufacturing emphasizes automation, informatization, and intelligence. Production monitoring images, as an important data source, can reflect the operating status of equipment, the appearance quality of products, and the production environment conditions. Production monitoring images are characterized by large data volume, strong real-time performance, and diversity.

[0039] Specifically, based on the needs of the production scenario, the specific content that needs to be monitored should be clearly defined, such as equipment operating status, product quality characteristics, or production environment parameters; the acquisition equipment should be installed in key locations on the production site to ensure coverage of all areas that need to be monitored; at the same time, the parameters of the acquisition system (such as exposure time, gain, frame rate, etc.) should be configured to adapt to different lighting conditions and production environments.

[0040] Furthermore, during the acquisition process, the acquired production monitoring images undergo preprocessing operations such as noise reduction, contrast enhancement, and distortion correction to improve image quality.

[0041] S2. The production monitoring image is divided into multiple non-overlapping blocks of a specified size using a block truncation algorithm based on absolute distance, and the bitmap, low quantization value, and high quantization value of each non-overlapping block are obtained. According to the preset key matrix and weight matrix, the secret information is embedded into the bitmap of each non-overlapping block using a block hiding algorithm to obtain the ciphertext bitmap of each non-overlapping block. The low quantization value, high quantization value, and ciphertext bitmap of each non-overlapping block are encoded, and the encoding result is encrypted and stored based on the NIST quantum-resistant cryptography standard algorithm.

[0042] It should be noted that image data needs to be secure, reliable, and traceable to prevent data leakage, tampering, or loss. Therefore, production monitoring images collected during intelligent manufacturing face multiple technical challenges: First, due to the massive amount of image data, direct storage or transmission would result in huge resource consumption, while lossless compression, although preserving all details, has low compression efficiency; second, traditional encryption methods (such as AES, RSA, etc.) may no longer be secure under the threat of quantum computing and cannot fully protect the confidentiality of image data; finally, to prevent images from being tampered with or forged, secret information needs to be embedded to verify their authenticity, but existing information hiding technologies are often independent of the compression and encryption processes, making efficient integrated processing difficult.

[0043] To address these issues, a comprehensive method is needed that can simultaneously achieve efficient lossy compression, post-quantum encryption, and information hiding to meet the security and practicality requirements of smart manufacturing for production monitoring images. Therefore, this embodiment proposes a comprehensive solution that combines post-quantum encryption technology, lossy compression algorithms, and information hiding technology.

[0044] The flowchart for step S2 is shown below. Figure 2 This includes steps S21 to S23, specifically:

[0045] S21. The production monitoring image is divided into multiple non-overlapping blocks of a specified size by a block truncation algorithm based on absolute distance, and the bitmap, low quantization value and high quantization value of each non-overlapping block are obtained.

[0046] It should be noted that the Absolute Moment Block Truncation Coding (AMBTC) algorithm is an image compression technique. Its core idea is to divide the image into several non-overlapping small blocks and achieve image compression by quantizing and simplifying the gray value distribution of the pixels in each block.

[0047] Specifically, the production monitoring image is divided into multiple non-overlapping blocks of a specified size using a block truncation algorithm based on absolute distance, and the bitmap, low quantization value, and high quantization value of each non-overlapping block are obtained. The specific process is as follows:

[0048] 1. Calculate the mean grayscale value of all pixels in the non-overlapping block.

[0049] 2. Classify pixels with gray values ​​greater than the mean into one category and write "1" in the corresponding position in the bitmap; classify pixels with gray values ​​not greater than the mean into another category and write "0" in the corresponding position in the bitmap; thus obtaining a bitmap of non-overlapping blocks.

[0050] 3. Calculate the mean gray value of all pixels of type I and take the rounded result as the high quantization value of the non-overlapping block; calculate the mean gray value of all pixels of type II and take the rounded result as the low quantization value of the non-overlapping block.

[0051] For example, for such Figure 4 The non-overlapping block shown has a mean grayscale value of 160.625 for all pixels. There are 10 pixels with grayscale values ​​greater than the mean, classified as Class 1 pixels, and a "1" is written in the corresponding position in the bitmap. There are 6 pixels with grayscale values ​​not greater than the mean, classified as Class 2 pixels, and a "0" is written in the corresponding position in the bitmap. Figure 4 The bitmap of the non-overlapping blocks shown is as follows Figure 5 As shown.

[0052] S22. Based on the preset key matrix and weight matrix, the secret information is embedded into the bitmap of each non-overlapping block using a block hiding algorithm to obtain the ciphertext bitmap of each non-overlapping block.

[0053] It should be noted that the block-based hiding algorithm is an information hiding algorithm for binary matrices. This algorithm divides a binary image into image blocks of the same size, introduces a key matrix and a weight matrix of the same size, and uses the block-based hiding algorithm to determine the difference based on the image block, the key matrix, the weight matrix and the secret information. Based on the difference, the information hiding method is determined, and the information is hidden in the image block according to the information hiding method, so that information hiding is performed once in each image block.

[0054] Specifically, based on the preset key matrix and weight matrix, the secret information is embedded into the bitmap of each non-overlapping block using a block-based hiding algorithm to obtain the ciphertext bitmap of each non-overlapping block. The specific process is as follows:

[0055] 1. Divide the non-overlapping bitmap into multiple sub-blocks of size 4×4.

[0056] 2. For any sub-block , pair of blocks The transformation is performed to obtain the transformed sub-block. The transformed sub-blocks are required. The following relationship must be satisfied:

[0057] ;

[0058] In the formula, , These are the key matrix and the weight matrix, respectively. This indicates summation. This represents the decimal number corresponding to the secret information. This represents the XOR operation. This indicates digit-wise multiplication. This indicates the modulo operation.

[0059] Among them, pair blocks Transformation refers to: transforming sub-blocks At least one element is swapped between 0 and 1, where swapping 0 and 1 means that when an element is equal to 1, it is set to 0; when an element is equal to 0, it is set to 1.

[0060] It should be noted that if the sub-block satisfy Then there is no need to modify the sub-blocks. To perform a transformation, that is, a sub-block It is itself a transformed sub-block .

[0061] In addition, when multiple transformation methods exist, the transformation method with the fewest transformations is selected for use on the sub-block. Transform the elements in the block to obtain the transformed sub-block. When multiple transformation methods with the fewest transformations exist, one transformation method is randomly selected from them to be used on the sub-block. Transform the elements in the block to obtain the transformed sub-block. The transformed sub-block obtained at this time Secret information has already been embedded in it.

[0062] The preset key matrix is ​​a 4×4 matrix, and the elements in the matrix are either 0 or 1; for example, a schematic diagram of the key matrix is ​​shown below. Figure 6 As shown.

[0063] The preset weight matrix is ​​a 4×4 matrix, and the elements in the matrix are integers in the range [1,7], with each integer in the range [1,7] appearing 2 or 3 times in the matrix; for example, a schematic diagram of the weight matrix is ​​shown below. Figure 7 As shown.

[0064] Furthermore, the secret information is extracted from the secret information sequence, and the length of the extracted secret information is equal to 3; the secret information sequence is a binary sequence composed of 0 and 1; in one embodiment, the secret information sequence can be randomly set; in another embodiment, the encoding result of key information in intelligent manufacturing can be used as the secret information sequence.

[0065] 3. The image block formed by arranging all the transformed sub-blocks in sequence is the ciphertext bitmap of the non-overlapping block. Thus, the secret information is embedded into the bitmap of the non-overlapping block.

[0066] For example, for such Figure 5 The bitmap shown contains only one sub-block. When the secret information is "101", the corresponding decimal number is... =5, pair of sub-blocks One transformation method is to transform the sub-blocks. The element in the first row and first column that is equal to 0 is set to 1, resulting in the transformed sub-block. like Figure 8 As shown in (1), at this time, the transformed sub-block Satisfying the relation: Another transformation method is to transform the sub-blocks. The element in the second row and third column that is equal to 1 is set to 0, resulting in the transformed sub-block. like Figure 8 As shown in (2), at this time, the transformed sub-block It also satisfies the following relation: In summary, regarding such Figure 5 The bitmap shown has two transformation methods, both with the same number of transformations (1). Therefore, one transformation method can be randomly selected to transform the sub-blocks. Transform the elements in the block to obtain the transformed sub-block. The transformed sub-block obtained at this time Secret information has already been embedded in it.

[0067] S23. Encode the low quantization value, high quantization value, and ciphertext bitmap of each non-overlapping block, encrypt the encoding result based on the NIST quantum-resistant cryptography standard algorithm, and store it.

[0068] Specifically, the low quantization value, high quantization value, and ciphertext bitmap of each non-overlapping block are encoded to obtain the encoding result of the production monitoring image.

[0069] For the low and high quantization values ​​of non-overlapping blocks: since both low and high quantization values ​​are obtained by averaging and rounding down the grayscale values ​​of pixels, the range of low and high quantization values ​​is [0, 255], a total of 256 values. Therefore, when encoding low and high quantization values ​​with a fixed length, the length of the encoding result is fixed to equal to... Specifically, the binary numbers of length 8 corresponding to the low quantization value and the high quantization value are respectively used as the encoding results of the low quantization value and the high quantization value, and the binary numbers are composed of 0 and 1.

[0070] For ciphertext bitmaps that do not overlap, since the ciphertext bitmap is essentially a matrix composed of 0s and 1s, it can be directly encrypted without encoding.

[0071] It should be noted that with the development of quantum computing technology, traditional public-key cryptography algorithms such as RSA, Diffie-Hellman, and elliptic curve cryptography are at risk of being broken by quantum computers. Post-quantum cryptography is a new generation of cryptographic algorithms that can resist attacks by quantum computers on existing cryptographic algorithms. Post-quantum cryptography is based on mathematical problems, and its core lies in using the computational complexity of certain mathematical problems to resist attacks by quantum computers.

[0072] Furthermore, encrypting the encoding results based on the NIST quantum-resistant cryptography standard algorithm can effectively resist quantum computing attacks and ensure the long-term security of the data.

[0073] Optionally, the data encryption process uses the NIST-standard CRYSTALS-Kyber algorithm (which is one of the algorithms included in the NIST quantum-resistant cryptography standard) to encapsulate the shared key. The shared key is used to symmetrically encrypt the low quantization value, high quantization value, and ciphertext bitmap encoding results of the non-overlapping blocks obtained after image segmentation, ensuring the quantum-resistant security of production monitoring images throughout the entire storage cycle.

[0074] Optionally, a hash algorithm (such as SM3) can be used to generate a digest of the image data, and the digest can be digitally signed using the CRYSTALS-Dilithium algorithm (another algorithm included in the NIST quantum-resistant cryptography standard) to achieve quantum-resistant integrity verification and watermark source tracing.

[0075] Among them, the NIST-standardized CRYSTALS-Kyber algorithm and CRYSTALS-Dilithium algorithm are well-known technologies and will not be elaborated here.

[0076] Regarding the specified size in step S21, in one embodiment, the specified size can be any pre-selected size.

[0077] Among them, the preselected size is equal to ,in, , All are integers in the range [1, 5], therefore , Different values ​​will result in different pre-selected sizes.

[0078] It should be noted that, in order to meet the requirements of compression efficiency, security protection and anti-tampering while minimizing the loss of image quality and information hiding performance, it is necessary to select an optimal block partitioning size; by calculating the comprehensive effect value of each pre-selected size, the size with the largest comprehensive effect value is selected as the specified size.

[0079] In another embodiment, the image compression degree, the loss caused by information hiding, and the loss caused by compression coding corresponding to each pre-selected size are weighted to obtain the comprehensive effect value of each pre-selected size; the pre-selected size with the largest comprehensive effect value is taken as the specified size.

[0080] The method for obtaining the comprehensive effect value of each pre-selected dimension in step S21 is as follows: Figure 3 This includes steps S211 to S214:

[0081] S211. Calculate the image compression degree corresponding to the pre-selected size based on the number of non-overlapping blocks corresponding to the pre-selected size and the amount of data stored for each non-overlapping block.

[0082] It's important to note that different block sizes result in varying degrees of image compression and image loss. Larger block sizes typically improve the compression ratio but may lead to greater image distortion; smaller block sizes preserve more detail but reduce compression efficiency. By determining the specified size that maximizes the overall effect, an optimal balance between compression efficiency and image quality can be found.

[0083] When encoding and compressing production monitoring images using a block truncation algorithm based on absolute distance, for each non-overlapping block corresponding to a pre-selected size, the information that needs to be stored includes the low quantization value, high quantization value, and bitmap of the non-overlapping block. Since both the low and high quantization values ​​are obtained by averaging and rounding the grayscale values ​​of pixels, and the grayscale value range is [0, 255], the low and high quantization values ​​have a total of 256 possible values. Therefore, when encoding the low and high quantization values ​​with a fixed length, the length of the encoded result is fixed to equal to... In other words, the amount of data storing the low-quantization value of each non-overlapping block is equal to 8, and the amount of data storing the high-quantization value of each non-overlapping block is equal to 8. Since a bitmap is essentially a matrix of 0s and 1s, it can be stored directly without encoding. Therefore, the amount of data storing each non-overlapping block of the bitmap is equal to the number of elements in the bitmap, i.e., the amount of data storing each non-overlapping block of the bitmap is equal to... In summary, the amount of data stored for each non-overlapping block is equal to... Pre-selected size The number of corresponding non-overlapping blocks is When encoding and compressing production monitoring images using a block truncation algorithm based on absolute distance, the amount of data used to encode and store the entire production monitoring image is equal to... .

[0084] If production monitoring images are directly encoded and compressed, the grayscale value of each pixel in the image needs to be encoded and compressed. Since the grayscale value ranges from [0, 255], a total of 256 values, when encoding grayscale values ​​with a fixed length, the length of the encoded result is fixed to equal to... In other words, the amount of data storing the grayscale value of each pixel is equal to 8, and the total number of pixels in the produced monitoring image is equal to In summary, when directly encoding and compressing production monitoring images, the amount of data required to encode and store the entire production monitoring image is equal to... .

[0085] The formula for calculating the degree of image compression corresponding to the pre-selected size is:

[0086] ;

[0087] Where, For pre-selected size, Indicates the pre-selected size The corresponding level of image compression, Indicates the pre-selected size The corresponding number of non-overlapping blocks, the amount of data stored in each non-overlapping block is equal to , To determine the size of the generated surveillance images.

[0088] It should be noted that when encoding and compressing production monitoring images using a block truncation algorithm based on absolute distance, the smaller the amount of data stored in the entire production monitoring image, the better the compression effect, and correspondingly, the greater the degree of image compression corresponding to the pre-selected size.

[0089] S212. Calculate the degree of loss caused by compression coding corresponding to the preselected size based on the difference between the gray values ​​of pixels in all non-overlapping blocks corresponding to the preselected size and the low quantization value and the high quantization value.

[0090] When encoding and compressing production monitoring images using a block truncation algorithm based on absolute distance, for each non-overlapping block corresponding to a pre-selected size, the decoding process involves replacing elements with equal "1" values ​​in the bitmap of the non-overlapping block with high quantization values ​​and replacing elements with equal "0" values ​​in the bitmap of the non-overlapping block with low quantization values. Therefore, the high and low quantization values ​​determine the degree of image quality loss of the decoded production monitoring image compared to the original production monitoring image: for a pixel in a non-overlapping block, if the element at the corresponding position in the bitmap is equal to "1", that is, for a type of pixel in the non-overlapping block, the degree of loss is equal to the difference between the grayscale value of the type of pixel and the high quantization value; for a pixel in a non-overlapping block, if the element at the corresponding position in the bitmap is equal to "0", that is, for a type of pixel in the non-overlapping block, the degree of loss caused by compression encoding is equal to the difference between the grayscale value of the type of pixel and the low quantization value.

[0091] The formula for calculating the degree of loss caused by compression encoding is:

[0092] ;

[0093] In the formula, For pre-selected size, Indicates the pre-selected size The degree of loss caused by the corresponding compression encoding. Indicates the pre-selected size The corresponding number of non-overlapping blocks, , The first The number of all Class I pixels and all Class II pixels in a non-overlapping block , The first The first non-overlapping block The first type of pixel and the first The grayscale values ​​of two types of pixels , The first High and low quantization values ​​of non-overlapping blocks To produce the size of the surveillance images, This represents the function that takes the absolute value.

[0094] It should be noted that, Indicates the first The first non-overlapping block The difference between the grayscale value and the high quantization value of a single pixel Indicates the first The first non-overlapping block The difference between the grayscale value and the low quantization value of each second-class pixel.

[0095] S213. Calculate the degree of loss caused by information hiding corresponding to the preselected size based on the number of secret information embeddings, the difference between low quantization value and high quantization value of all non-overlapping blocks corresponding to the preselected size.

[0096] It should be noted that factors such as the number of times the secret information is embedded and the difference between low and high quantization values ​​can affect the degree of loss caused by information hiding. Choosing an appropriate block size can reduce the impact of information hiding on image quality, while ensuring the robustness and invisibility of the watermark information.

[0097] The secret information is embedded into the bitmaps of each non-overlapping block using a block-based hiding algorithm. To obtain the ciphertext bitmap of each non-overlapping block: first, the bitmap of the non-overlapping block is divided into multiple sub-blocks of size 4×4. Then, each sub-block is of size 4×4. The bitmap of non-overlapping blocks is divided into Each sub-block is transformed by setting elements equal to 1 to 0 and elements equal to 0 to 1. This results in a transformed sub-block. During decoding, elements originally replaced by high quantization values ​​are replaced by low quantization values, or vice versa. Regardless of the cause, the decoded production monitoring image will suffer a quality loss compared to the original image, with the loss in each sub-block equal to the difference between the high and low quantization values. Then the loss caused by information hiding for each non-overlapping block is equal to .

[0098] The formula for calculating the degree of loss caused by information hiding is:

[0099] ;

[0100] In the formula, For pre-selected size, Indicates the pre-selected size The extent of loss caused by the corresponding information hiding, Indicates the pre-selected size The number of corresponding non-overlapping blocks, and the number of times the secret information of each non-overlapping block corresponding to the pre-selected size is embedded equal to... , , Pre-selected size The corresponding number High and low quantization values ​​of non-overlapping blocks To determine the size of the generated surveillance images.

[0101] S214. Weight the degree of image compression, the degree of loss caused by information hiding, and the degree of loss caused by compression coding for each pre-selected size to obtain the comprehensive effect value of each pre-selected size.

[0102] It should be noted that the overall performance value considers multiple factors, including the degree of image compression, the loss caused by compression coding, and the loss caused by information hiding, and is calculated through weighted averages. Therefore, selecting the specified size with the largest overall performance value ensures that the entire method achieves optimal performance in compression, encryption, and watermark embedding.

[0103] The formula for calculating the overall effect value of each pre-selected size is as follows:

[0104] ;

[0105] Where, Indicates the pre-selected size The overall effect value, Indicates the pre-selected size The corresponding level of image compression, Indicates the pre-selected size The degree of loss caused by the corresponding compression encoding. Indicates the pre-selected size The extent of loss caused by the corresponding information hiding, Indicates the loss threshold. This represents the natural exponential function.

[0106] in, , These are the first weight and the second weight, respectively. , All are greater than 0, and The specific values ​​of the first and second weights can be set according to the actual application scenario and requirements. In this invention, the first weight is preset to 0.4 and the second weight is preset to 0.6.

[0107] The specific value of the loss threshold can be set according to the actual application scenario and needs, and the range of the loss threshold is [5,20]. In this invention, the loss threshold is set to 13.

[0108] It should be noted that the calculation of the overall effect value includes the image compression degree as an indicator. Selecting the specified size with the largest overall effect value can ensure that the highest possible compression ratio is achieved while meeting certain image quality requirements, thereby effectively reducing storage and transmission costs. In addition, by analyzing the degree of loss caused by compression coding and the degree of loss caused by information hiding, the overall effect value can comprehensively reflect the impact of different block partitioning sizes on image quality. Selecting the specified size with the largest overall effect value can minimize image distortion and visual quality degradation while ensuring compression efficiency.

Claims

1. A data security protection method for intelligent manufacturing based on post-quantum technology, characterized in that, include: The production monitoring image is divided into multiple non-overlapping blocks of a specified size using a block truncation algorithm based on absolute distance, and the bitmap, low quantization value, and high quantization value of each non-overlapping block are obtained. According to the preset key matrix and weight matrix, the secret information is embedded into the bitmap of each non-overlapping block using a block hiding algorithm. The ciphertext bitmap, low quantization value, and high quantization value of each non-overlapping block are encoded, and the encoding result is encrypted and stored based on the NIST quantum-resistant cryptography standard algorithm. The specified size is the pre-selected size with the highest overall effect value. The method for obtaining the overall effect value of each pre-selected size is as follows: The image compression degree corresponding to the pre-selected size is calculated based on the number of non-overlapping blocks corresponding to the pre-selected size and the amount of data stored for each non-overlapping block. The loss caused by compression coding corresponding to the pre-selected size is calculated based on the difference between the grayscale value of the pixel in all non-overlapping blocks corresponding to the pre-selected size and the low quantization value and the high quantization value. The loss caused by information hiding corresponding to the pre-selected size is calculated based on the number of times the secret information is embedded in all non-overlapping blocks corresponding to the pre-selected size and the difference between the low quantization value and the high quantization value. The image compression degree, the loss caused by information hiding, and the loss caused by compression coding corresponding to each pre-selected size are weighted to obtain the comprehensive effect value of each pre-selected size.

2. The intelligent manufacturing data security protection method based on post-quantum technology according to claim 1, characterized in that, Obtaining the bitmap, low quantization value, and high quantization value of each non-overlapping block includes: Calculate the mean grayscale value of all pixels in the non-overlapping block; Pixels with gray values ​​greater than the mean are classified into one class of pixels, and "1" is written in the corresponding position in the bitmap; pixels with gray values ​​not greater than the mean are classified into two classes of pixels, and "0" is written in the corresponding position in the bitmap; this is how to obtain a bitmap of non-overlapping blocks. Calculate the mean grayscale value of all pixels of class 1, and use the rounded result as the high quantization value of the non-overlapping block; calculate the mean grayscale value of all pixels of class 2, and use the rounded result as the low quantization value of the non-overlapping block.

3. The intelligent manufacturing data security protection method based on post-quantum technology according to claim 1, characterized in that, The preset key matrix is ​​a 4×4 matrix, and the elements in the matrix are either 0 or 1.

4. The intelligent manufacturing data security protection method based on post-quantum technology according to claim 1, characterized in that, The weight matrix is ​​a 4×4 matrix, and the elements in the matrix are integers in the range [1,7], and each integer in the range [1,7] appears 2 or 3 times in the matrix.

5. The intelligent manufacturing data security protection method based on post-quantum technology according to claim 1, characterized in that, The step of embedding secret information into the bitmap of each non-overlapping block using a block-based hiding algorithm includes: Divide the non-overlapping bitmap into multiple sub-blocks of size 4×4; for any sub-block , pair of blocks Perform the transformation to obtain the transformed sub-block. The transformed sub-blocks are required. satisfy In the formula, , These are the key matrix and the weight matrix, respectively. This indicates summation. This represents the decimal number corresponding to the secret information. This represents the XOR operation. This indicates digit-wise multiplication. This represents the modulo operation; The pair of blocks Transformation refers to: transforming sub-blocks At least one element is swapped between 0 and 1.

6. The intelligent manufacturing data security protection method based on post-quantum technology according to claim 1, characterized in that, The preselected size equals ,in, , All are integers in the range [1, 5].

7. The intelligent manufacturing data security protection method based on post-quantum technology according to claim 6, characterized in that, The calculation of the image compression degree corresponding to the pre-selected size includes: ； In the formula, For pre-selected size, Indicates the pre-selected size The corresponding level of image compression, Indicates the pre-selected size The corresponding number of non-overlapping blocks, the amount of data stored in each non-overlapping block is equal to , To determine the size of the generated surveillance images.

8. The intelligent manufacturing data security protection method based on post-quantum technology according to claim 6, characterized in that, The calculation of the loss caused by compression encoding corresponding to the pre-selected size includes: ； In the formula, For pre-selected size, Indicates the pre-selected size The degree of loss caused by the corresponding compression encoding. Indicates the pre-selected size The corresponding number of non-overlapping blocks, , Respectively The number of all Class I pixels and all Class II pixels in a non-overlapping block , Respectively The first non-overlapping block The first type of pixel and the first The grayscale values ​​of two types of pixels , Respectively High and low quantization values ​​of non-overlapping blocks To produce the size of the surveillance images, This represents the function that takes the absolute value.

9. The intelligent manufacturing data security protection method based on post-quantum technology according to claim 6, characterized in that, The calculation of the degree of loss caused by information hiding corresponding to the pre-selected size includes: ； In the formula, For pre-selected size, Indicates the pre-selected size The extent of loss caused by the corresponding information hiding, Indicates the pre-selected size The number of corresponding non-overlapping blocks, and the number of times the secret information of each non-overlapping block corresponding to the pre-selected size is embedded equal to... , , Pre-selected size The corresponding number High and low quantization values ​​of non-overlapping blocks To determine the size of the generated surveillance images.

10. The intelligent manufacturing data security protection method based on post-quantum technology according to claim 6, characterized in that, The step of weighting the image compression degree, the loss due to information hiding, and the loss due to compression coding for each pre-selected size to obtain a comprehensive effect value for each pre-selected size includes: ； In the formula, Indicates the pre-selected size The overall effect value, , , These represent the pre-selected dimensions. The corresponding image compression level, the degree of loss caused by compression coding, and the degree of loss caused by information hiding. Indicates the loss threshold. This represents the natural exponential function. , These are the first weight and the second weight, respectively. , are greater than 0, and .

Images