AID authentication anti-counterfeiting method based on digital block chain

Abstract

The invention relates to the technical field of artwork authentication and anti-counterfeiting, in particular to an AID authentication and anti-counterfeiting method based on a digital block chain, and the method comprises the steps: S1, collecting the attribute data of artworks; s2, preprocessing the image data according to a preprocessing method to obtain preprocessed image data; s3, performing feature processing on the preprocessed image data to obtain feature image data; s4, performing hash processing on the feature image data and the non-image data to generate a hash value, and storing the hash value and the artwork attribute data on a block chain to generate a digital authentication certificate; s5, constructing an identification code conversion model, and converting the hash value into an artwork identification code through the identification code conversion model; and S6, outputting the user operation process according to the security processing method. The safety of artwork circulation is improved.

Description

Technical Field

[0001] The present invention relates to the field of artwork authentication and anti-counterfeiting technology, and in particular to an AID authentication and anti-counterfeiting method based on digital blockchain. Background Art

[0002] With the development of digitalization and globalization, the problem of counterfeiting is becoming increasingly serious, resulting in huge economic losses and a crisis of trust. In order to meet these challenges, an efficient and reliable anti-counterfeiting technology is urgently needed. Blockchain is a distributed ledger technology with the characteristics of decentralization, immutability and transparency, making it an ideal tool for anti-counterfeiting and authentication, because once any information is recorded on the blockchain, it cannot be tampered with. Artificial intelligence technology performs well in data analysis and pattern recognition. Through artificial intelligence, intelligent analysis and verification of anti-counterfeiting information can be achieved, and the automation and accuracy of the anti-counterfeiting system can be improved. The development of big data technology makes it possible to collect and analyze large amounts of data, which can be used to train artificial intelligence models to improve their recognition and verification capabilities.

[0003] Chinese patent publication number: CN107341673A discloses a method for authenticating artworks, including the following steps: 1. Identity registration: Art institutions, artists and collectors register their identities on the server; 2. Artwork filing: Upload the full picture of the artwork and its detailed microscopic picture to the server, and fill in the relevant information of the artwork; Review the uploaded artwork, and automatically generate an electronic filing certificate after the review is passed; 3. The artwork that has been filed is visually identified by the artist himself. If it is confirmed that the artwork is created by the artist himself, the artwork authentication certificate is automatically generated after uploading the certification materials. However, this solution is difficult to protect the data security during the circulation of artworks, and it is easy to illegally obtain artwork information, and it is difficult to ensure the security of artwork circulation. Summary of the invention

[0004] To this end, the present invention provides an AID authentication and anti-counterfeiting method based on digital blockchain to overcome the problem of low security in the circulation of artworks in the prior art.

[0005] To achieve the above-mentioned purpose, the present invention provides an AID authentication and anti-counterfeiting method based on digital blockchain, comprising:

[0006] Step S1, collecting artwork attribute data;

[0007] Step S2, preprocessing the image data according to a preprocessing method to obtain preprocessed image data;

[0008] Step S3, performing feature processing on the preprocessed image data to obtain feature image data;

[0009] Step S4, performing hash processing on the feature image data and non-image data to generate a hash value, and storing the hash value and artwork attribute data on the blockchain to generate a digital authentication certificate;

[0010] Step S5, constructing an identification code conversion model, and converting the hash value into an artwork identification code through the identification code conversion model;

[0011] Step S6: output the user operation process according to the security processing method.

[0012] Furthermore, the pretreatment method comprises:

[0013] Step S200, resizing and grayscale processing the image data to obtain standard image data;

[0014] Step S201, performing denoising processing on the standard image data by Gaussian filtering to obtain denoised image data;

[0015] Step S202: performing edge processing on the denoised image data to obtain pre-processed image data.

[0016] Furthermore, in the step S200, when the image data is resized and grayscale processed, the image data is resized to a preset size through an image editing tool, and the image data is converted into a grayscale image, and the converted image data is output as standard image data;

[0017] In step S201, when the standard image data is subjected to denoising processing by Gaussian filtering to obtain denoised image data, the Gaussian filtering calculation formula is set as:

[0018]

[0019] Where G(x,y) is the value of the Gaussian function at the point (x,y);

[0020] x is the distance in the horizontal direction;

[0021] y is the distance in the vertical direction;

[0022] π is the ratio of a circle to a circle;

[0023] e is the base of natural logarithm;

[0024] σ is the standard deviation of the Gaussian function, and its calculation formula is:

[0025]

[0026] Where W is the width of the image;

[0027] H is the height of the image;

[0028] In the step S202, when edge processing is performed on the denoised image data, edge processing is performed on the denoised image data by using a Laplace operator to obtain pre-processed image data.

[0029] Furthermore, in step S3, when the preprocessed image data is feature processed, the preprocessed image data is feature processed by Hough transform, each edge point in the image is mapped to a straight line in the Hough space by using Hough transform, an intersection is found in the Hough space, the accumulation degree Q of the intersection is compared with the preset accumulation degree Q0, the existence of the straight line is judged according to the comparison result, and the judgment result is output, wherein:

[0030] When Q<Q0, it is determined that the existence of the straight line does not meet the standard and the straight line is not output;

[0031] When Q≥Q0, the existence of the straight line is determined to be satisfactory, and the parameters of the straight line are output as feature image data.

[0032] Furthermore, in step S4, when the feature image data and non-image data are hashed to generate a hash value, and the hash value and artwork attribute data are stored on the blockchain to form an electronic certificate, the feature image data and non-image data are hashed using SHA-256 in the hash function to generate a hash value, and the hash value, feature image data and non-image data are stored in a preset smart contract, and a digital authentication certificate containing the feature image data, non-image data, hash value, transaction ID and contract address on the blockchain is generated.

[0033] Furthermore, in step S5, when constructing the identification code conversion model, the artwork identification code conversion database is divided into a conversion training set of 70%, a conversion verification set of 20%, and a conversion test set of 10%. The convolutional neural network model is trained by the conversion verification set, the conversion verification set is input into the trained convolutional neural network model for parameter optimization, and the conversion test set is input into the convolutional neural network model after parameter optimization for testing to obtain the conversion accuracy E, the conversion accuracy E is compared with the preset conversion accuracy E0, the training compliance of the convolutional neural network model is judged according to the comparison result, and the judgment result is output, wherein:

[0034] When E≥E0, the convolutional neural network model is judged to have met the training standards, the convolutional neural network model is output as an identification code conversion model, and the hash value is converted into an artwork identification code;

[0035] When E<E0, it is determined that the training of the convolutional neural network model does not meet the standards, and a reminder to update the artwork identification code conversion database is pushed to the administrator, who updates the artwork identification code conversion database and repeats step S5 with the updated artwork identification code conversion database until the training of the convolutional neural network model meets the standards.

[0036] Furthermore, the safety processing method includes:

[0037] Step S601, obtaining the user requested target and outputting it according to the user requested target;

[0038] Step S602, correcting the process of outputting the target according to the user's request through the user's login environment data and the login environment recognition model;

[0039] Step S603, obtaining the number of times the user login environment security does not meet the standard within a preset identification period, comparing the number of times the user does not meet the standard with the preset number of times the user does not meet the standard, judging the user's frequency of not meeting the standard based on the comparison result, and outputting the judgment result.

[0040] Furthermore, in step S601, the user request target is obtained and outputted according to the user request target, wherein:

[0041] When the user's request target is to authenticate a work of art, a work of art attribute data collection window is pushed to the user, and the user enters the work of art attribute data in the work of art attribute data collection window, and steps S2 to S5 are executed according to the data in the work of art attribute data collection window;

[0042] When the user's request target is to query artworks, obtain the artwork identification code input by the user, and provide the user with a digital authentication certificate corresponding to the artwork identification code input by the user;

[0043] When the user's request target is to trade artworks, the artwork identification code entered by the user is obtained, and output is performed according to the artwork identification code entered by the user, where:

[0044] If the artwork identification code input by the user exists, the artwork attribute data corresponding to the artwork identification code input by the user is provided to the user, and a transaction warning window is pushed to the transaction administrator, who monitors the process of the artwork transaction and obtains the artwork attribute data after the transaction, and executes steps S2 to S5 according to the artwork attribute data after the transaction;

[0045] If the artwork identification code entered by the user does not exist, blank data is provided to the user.

[0046] Furthermore, in step S602, the user login environment data is acquired, the user login environment data is input into the login environment recognition model for recognition, the recognition result is output, the security of the user login environment is judged according to the recognition result, and the process of outputting according to the user request target is corrected according to the judgment result, wherein:

[0047] When the recognition result of the login environment recognition model is that there is no potential safety hazard, it is determined that the security of the user login environment meets the standard, and the process of outputting according to the user request target is not modified;

[0048] When the recognition result of the login environment recognition model is that there is a security risk, it is determined that the security of the user login environment does not meet the standard, and the process of outputting according to the user's request target is corrected to output blank data.

[0049] Further, in the step S603, the number of times P of the user login environment security not meeting the standard within the preset identification period is obtained, the number of times P not meeting the standard is compared with the preset number of times P0 not meeting the standard, the frequency of user not meeting the standard is judged according to the comparison result, and the correction process is optimized according to the judgment result, wherein:

[0050] When P≤P, the user is judged as a low-frequency user that does not meet the standard, and the correction process is not optimized;

[0051] When P>P0, the user is determined to be a high-frequency non-compliant user, the user is marked as a dangerous user, and the recognition result of the login environment recognition model is optimized to have a security risk.

[0052] Compared with the prior art, the beneficial effects of the present invention are that the method collects artwork attribute data through step S1 to ensure that the basic information of the digital authentication certificate is accurate and provides a reliable basis for subsequent processing; the method preprocesses the image data according to the preprocessing method through step S2 to improve image quality, reduce noise and interference, and lay a foundation for subsequent feature extraction and hash processing; the method performs feature processing on the preprocessed image data through step S3, thereby extracting unique features of the artwork and enhancing the uniqueness and anti-counterfeiting ability of the digital authentication certificate; the method performs hash processing on the feature image data and non-image data through step S4 to generate a hash value, and stores the hash value and artwork attribute data on the blockchain to utilize the irreversibility of the hash algorithm and the non-tamperability of the blockchain to ensure the security and credibility of the digital authentication certificate; the method converts the complex hash value into an artwork identification code that is easy to identify and remember through step S5, thereby facilitating user operation and verification; the method outputs the user operation process according to the security processing method through step S6 to ensure the compliance and security of the artwork authentication process and improve user experience and satisfaction.

[0053] In particular, in step S200, the image data is adjusted to a uniform size and grayscale format, which is conducive to the standardization and automation of subsequent processing and improves the reliability and stability of the overall process, while the grayscale processing can remove color interference and make the texture, edge and other features in the image more prominent, which is conducive to subsequent feature extraction and recognition;

[0054] In step S201, the image data is de-noised by using a unique weighted average method of Gaussian filtering, which has the effects of smoothing the image, reducing noise and retaining edges;

[0055] In step S202, the denoised image data is subjected to edge processing by the Laplace operator, so as to obtain pre-processed image data with prominent edge information and reduced redundant information, which is of great significance for subsequent image processing and analysis tasks.

[0056] In particular, in step S3, the straight line parameters obtained by Hough transform are quantified and comparable, which is convenient for subsequent image processing and analysis, and the features such as straight lines in the image are extracted in a parameterized form, which can greatly simplify the subsequent image recognition, classification and matching tasks, and improve processing efficiency and accuracy.

[0057] In particular, in step S4, the SHA-256 hash function ensures the integrity of the data. Any modification to the data will result in a change in the hash value, which will be recognized and rejected by the nodes in the blockchain network. At the same time, the distributed and decentralized nature of the blockchain makes the data stored on it highly tamper-proof, and each transaction on the blockchain is assigned a unique transaction ID and associated with the smart contract address. This allows the circulation history of the artwork to be clearly traced and verified, and the automatic execution characteristics of the smart contract reduce the need for manual intervention and improve the efficiency and accuracy of data processing.

[0058] In particular, in step S5, a high accuracy rate of the identification code conversion model is ensured through a rigorous training, verification and testing process, thereby improving the reliability of the artwork identification code conversion. At the same time, the use of independent verification sets and test sets to train and evaluate the model helps to enhance the generalization ability of the model, enabling it to better adapt to new data and scenarios. Moreover, when the database needs to be updated, the update process can be easily triggered and the model can be retrained to ensure the flexibility and scalability of the model, thereby realizing the automation and intelligence of model training, reducing the need for manual intervention, and improving work efficiency and accuracy.

[0059] In particular, in step S601, an intuitive and easy-to-use user interface and interaction method are provided, so that users can easily complete operations such as artwork authentication, transactions and inquiries. Through automated and digital management processes, the system improves the efficiency of artwork authentication, transactions and inquiries, reduces human errors and costs, and at the same time, transaction monitoring and artwork identification code verification mechanisms ensure the security and reliability of the transaction process, reducing the risk of fraud and illegal transactions.

[0060] In particular, in step S602, by introducing the login environment recognition model, potential security threats can be identified more accurately and corresponding measures can be taken to prevent them, thereby significantly improving the overall security. When security risks are found in the user login environment, the output of sensitive information can be automatically stopped, effectively protecting the user's personal privacy and rights.

[0061] In particular, in step S603, by identifying and strengthening monitoring of high-frequency non-compliant users, potential security threats can be more effectively prevented and overall security can be improved. At the same time, through dynamic risk assessment and personalized security policies, the system can more accurately judge the security of the user's login environment and reduce misjudgments and missed judgments. For low-frequency non-compliant users, the system maintains regular monitoring and does not affect their normal usage experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 This is a flow chart of the AID authentication and anti-counterfeiting method based on digital blockchain in this embodiment;

[0063] Figure 2 Schematic diagram of the pretreatment method of this embodiment;

[0064] Figure 3 Schematic diagram of the process of the security processing method of this embodiment. DETAILED DESCRIPTION

[0065] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0066] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0067] See also Figure 1 As shown, it is a flow chart of the AID authentication and anti-counterfeiting method based on digital blockchain in this example, and the method includes:

[0068] Step S1, collecting artwork attribute data;

[0069] Step S2, preprocessing the image data according to a preprocessing method to obtain preprocessed image data;

[0070] Step S3, performing feature processing on the preprocessed image data to obtain feature image data;

[0071] Step S4, performing hash processing on the feature image data and non-image data to generate a hash value, and storing the hash value and artwork attribute data on the blockchain to generate a digital authentication certificate;

[0072] Step S5, constructing an identification code conversion model, and converting the hash value into an artwork identification code through the identification code conversion model;

[0073] Step S6: output the user operation process according to the security processing method.

[0074] Specifically, the method is applied to the field of artwork authentication and anti-counterfeiting. It provides basic information for subsequent digital authentication by collecting artwork attribute data, preprocessing image data and feature processing, and converts the hash value into an artwork identification code that is easy to identify and remember through hash processing and identification code conversion. It also monitors and records the user operation process according to the security processing method to ensure the compliance and security of the artwork authentication process. In particular, the method collects artwork attribute data through step S1 to ensure that the basic information of the digital authentication certificate is accurate, providing a reliable basis for subsequent processing. The method preprocesses the image data according to the preprocessing method through step S2 to improve image quality, reduce noise and interference, and lay the foundation for subsequent feature extraction and hash processing. The method performs feature processing on the pre-processed image data in step S3, thereby extracting the unique features of the artwork and enhancing the uniqueness and anti-counterfeiting ability of the digital authentication certificate. The method performs hash processing on the feature image data and non-image data in step S4 to generate a hash value, and stores the hash value and artwork attribute data on the blockchain to utilize the irreversibility of the hash algorithm and the non-tamperability of the blockchain to ensure the security and credibility of the digital authentication certificate. The method converts the complex hash value into an artwork identification code that is easy to identify and remember in step S5, thereby facilitating user operation and verification. The method outputs the user operation process according to the security processing method in step S6 to ensure the compliance and security of the artwork authentication process and improve user experience and satisfaction.

[0075] Specifically, in step S1, the artwork attribute data includes image data and non-image data, wherein the non-image data includes the name of the artwork, the name of the creator of the artwork, the creation time of the artwork, the size, weight, material, color, shape, structure, pattern, the origin of the artwork, ownership history, exhibition history, restoration record, valuation, transaction history and copyright, and the image data refers to digital data that describes and stores the visual information of the artwork. This embodiment does not limit the method for collecting the artwork attribute data. Relevant technical personnel in this field can freely set it according to actual needs, as long as the demand for collecting the artwork attribute data is met. For example, the method for collecting the artwork attribute data can be set to capture the image of the artwork with a high-resolution camera, perform three-dimensional modeling and internal structure scanning of the artwork through a 3D scanner, laser scanner and X-ray, and fill in the non-image data by an administrator.

[0076] See also Figure 2 As shown, it is a schematic diagram of the process of the pretreatment method of this embodiment, and the pretreatment method includes:

[0077] Step S200, resizing and grayscale processing the image data to obtain standard image data;

[0078] Step S201, performing denoising processing on the standard image data by Gaussian filtering to obtain denoised image data;

[0079] Step S202: performing edge processing on the denoised image data to obtain pre-processed image data.

[0080] Specifically, in step S200, when the image data is resized and gray-scaled, the image data is resized to a preset size through an image editing tool, and the image data is converted into a gray-scale image, and the converted image data is output as standard image data.

[0081] Specifically, the image editing tool refers to software and applications used to process and modify images. This embodiment does not limit the specific type of image editing tools. Relevant technical personnel in this field can freely set them according to actual needs. It only needs to meet the need to adjust the image data in the artwork attribute data. For example, the image editing tool can be set to Adobe image processing software. The preset size refers to the pre-set image width and height. This embodiment does not limit the value of the preset size. Relevant technical personnel in this field can freely set them according to actual needs. It only needs to meet the need for uniform image data size. For example, the preset size can be set to 3000 pixels in width and 4000 pixels in height. The grayscale processing refers to the process of converting a color image into a grayscale image. The grayscale image refers to an image with only one sampled color per pixel.

[0082] Specifically, in step S200, the image data is adjusted to a uniform size and grayscale format, which is conducive to the standardization and automation of subsequent processing and improves the reliability and stability of the overall process. Grayscale processing can remove color interference and make the texture, edge and other features in the image more prominent, which is conducive to subsequent feature extraction and recognition.

[0083] Specifically, in step S201, when the standard image data is subjected to denoising processing by Gaussian filtering to obtain denoised image data, the Gaussian filtering calculation formula is set as:

[0084]

[0085] Where G(x,y) is the value of the Gaussian function at the point (x,y);

[0086] x is the distance in the horizontal direction;

[0087] y is the distance in the vertical direction;

[0088] π is the ratio of a circle to a circle;

[0089] e is the base of natural logarithm;

[0090] σ is the standard deviation of the Gaussian function, and its calculation formula is:

[0091]

[0092] Where W is the width of the image;

[0093] H is the height of the image.

[0094] Specifically, the Gaussian filter refers to a smoothing filter technology used for image processing, the denoising process refers to the process of reducing noise in an image and improving image quality, the center point refers to the center position of the Gaussian filter, and its coordinates are (0, 0), the point (x, y) refers to each pixel point in the image, the horizontal distance refers to the distance from the point (x, y) to the center point of the Gaussian filter in the horizontal direction, the vertical distance refers to the distance from the point (x, y) to the center point of the Gaussian filter in the vertical direction, the width of the image refers to the number of pixels of the image in the horizontal direction, and the height of the image refers to the number of pixels of the image in the vertical direction.

[0095] Specifically, in step S201, the denoising process of the image data is implemented through the unique weighted averaging method of Gaussian filtering, which has the effects of smoothing the image, reducing noise and retaining edges.

[0096] Specifically, in step S202, when edge processing is performed on the denoised image data, edge processing is performed on the denoised image data by using a Laplace operator to obtain pre-processed image data.

[0097] Specifically, the edge processing refers to the process of identifying areas in an image that meet a preset brightness change, and the Laplace operator refers to a second-order differential operator used to measure the second-order change of pixel grayscale values ​​in an image.

[0098] Specifically, in step S202, the denoised image data is subjected to edge processing by the Laplace operator, so that preprocessed image data with prominent edge information and reduced redundant information can be obtained, which is of great significance for subsequent image processing and analysis tasks.

[0099] Specifically, in step S3, when the pre-processed image data is feature processed, the pre-processed image data is feature processed by Hough transform, each edge point in the image is mapped to a straight line in the Hough space by using Hough transform, and the intersection point is found in the Hough space, and the accumulation degree Q of the intersection point is compared with the preset accumulation degree Q0, and the existence of the straight line is judged according to the comparison result, and output according to the judgment result, wherein:

[0100] When Q<Q0, it is determined that the existence of the straight line does not meet the standard and the straight line is not output;

[0101] When Q≥Q0, the existence of the straight line is determined to be satisfactory, and the parameters of the straight line are output as feature image data.

[0102] Specifically, the Hough transform refers to a technology used to detect simple shapes in an image, the feature processing refers to using the Hough transform to extract straight line features from preprocessed image data, the edge point refers to a point in the image where brightness and color change significantly, the Hough space refers to a two-dimensional space used to represent the parameters of a straight line in an image, the intersection point refers to the intersection of multiple straight lines mapped from edge points in the image, the accumulation degree Q of the intersection point refers to the number of straight lines passing through the intersection point, the preset accumulation degree Q0 refers to a preset value used to reflect the existence of a straight line, and this embodiment does not limit the specific numerical value of the preset accumulation degree Q0. Relevant technical personnel in this field can freely set it according to actual needs, as long as the need to judge the existence of a straight line is met. For example, the preset accumulation degree Q0 can be set to 3, and the parameters of the straight line refer to numerical values ​​that describe the characteristics of the straight line in the image, such as the slope, intercept, distance and angle of the straight line.

[0103] Specifically, in step S3, the straight line parameters obtained by Hough transform are quantified and comparable, which is convenient for subsequent image processing and analysis, and the features such as straight lines in the image are extracted in a parameterized form, which can greatly simplify subsequent image recognition, classification and matching tasks, and improve processing efficiency and accuracy.

[0104] Specifically, in step S4, when the feature image data and non-image data are hashed to generate a hash value, and the hash value and artwork attribute data are stored on the blockchain to form an electronic certificate, the feature image data and non-image data are hashed using SHA-256 in the hash function to generate a hash value, and the hash value, feature image data and non-image data are stored in a preset smart contract, and a digital authentication certificate containing the feature image data, non-image data, hash value, transaction ID and contract address on the blockchain is generated.

[0105] Specifically, the hash function refers to a function that maps input data of arbitrary length to an output of fixed length, the hash processing refers to the process of calculating a hash value of a preset length from pre-processed attribute data through a hash function, the SHA-256 refers to an encrypted hash function that converts data input of arbitrary length into a fixed-length 256-bit hash value, the blockchain refers to distributed ledger technology, the hash value refers to an output value of a fixed length generated after hashing the input data through a hash function, the preset smart contract refers to a program that is pre-written and deployed on the blockchain and runs automatically on the blockchain, the transaction ID on the blockchain refers to an identifier that uniquely identifies a transaction in the blockchain network, the contract address refers to an address that uniquely identifies a smart contract in the blockchain network, and the digital authentication certificate refers to a digitized certification document.

[0106] Specifically, in step S4, the SHA-256 hash function ensures the integrity of the data. Any modification to the data will result in a change in the hash value, which will be recognized and rejected by the nodes in the blockchain network. At the same time, the distributed and decentralized nature of the blockchain makes the data stored on it highly tamper-proof, and each transaction on the blockchain is assigned a unique transaction ID and associated with the smart contract address. This allows the circulation history of the artwork to be clearly traced and verified, and the automatic execution characteristics of the smart contract reduce the need for manual intervention and improve the efficiency and accuracy of data processing.

[0107] Specifically, in step S5, when constructing the identification code conversion model, the artwork identification code conversion database is divided into a 70% conversion training set, a 20% conversion verification set, and a 10% conversion test set. The convolutional neural network model is trained by the conversion verification set, the conversion verification set is input into the trained convolutional neural network model for parameter optimization, and the conversion test set is input into the convolutional neural network model after parameter optimization for testing to obtain the conversion accuracy E, the conversion accuracy E is compared with the preset conversion accuracy E0, the training compliance of the convolutional neural network model is judged according to the comparison result, and the judgment result is output, wherein:

[0108] When E≥E0, the convolutional neural network model is judged to have met the training standards, the convolutional neural network model is output as an identification code conversion model, and the hash value is converted into an artwork identification code;

[0109] When E<E0, it is determined that the training of the convolutional neural network model does not meet the standards, and a reminder to update the artwork identification code conversion database is pushed to the administrator, who updates the artwork identification code conversion database and repeats step S5 with the updated artwork identification code conversion database until the training of the convolutional neural network model meets the standards.

[0110] Specifically, the artwork identification code conversion database refers to a database containing hash values ​​and artwork identification codes corresponding to the hash values, the conversion training set refers to a data set used to train a convolutional neural network model, the conversion verification set refers to a data set used to verify the performance of the convolutional neural network model and optimize parameters during the training of the convolutional neural network model, the conversion test set refers to a data set used to test the performance of the convolutional neural network model after the training is completed, the convolutional neural network model refers to a deep learning model used to process data, the parameter optimization refers to adjusting the parameters of the convolutional neural network model during the training process to minimize the loss function of the model on the verification set, thereby improving the performance of the model on unseen data, the conversion accuracy E refers to the performance index of the model on the test set, indicating the proportion of the model correctly converting the hash value into the corresponding artwork identification code, the preset conversion accuracy E0 refers to a preset value used to judge whether the convolutional neural network model has been trained to meet the standards, this embodiment does not limit the value of the preset conversion accuracy, and relevant technical personnel in the field can freely set it according to actual needs, as long as the need to judge whether the convolutional neural network model has been trained to meet the standards is met, such as setting the preset conversion accuracy E0 = 0.99, the administrator refers to the person responsible for managing and maintaining the artwork identification code conversion database.

[0111] Specifically, in step S5, a high accuracy rate of the identification code conversion model is ensured through a rigorous training, verification and testing process, thereby improving the reliability of the artwork identification code conversion. At the same time, the use of independent verification sets and test sets to train and evaluate the model helps to enhance the generalization ability of the model, enabling it to better adapt to new data and scenarios. Moreover, when the database needs to be updated, the update process can be easily triggered and the model can be retrained to ensure the flexibility and scalability of the model, thereby realizing the automation and intelligence of model training, reducing the need for manual intervention, and improving work efficiency and accuracy.

[0112] See also Figure 3 As shown, it is a schematic diagram of the process of the security processing method of this embodiment, and the security processing method includes:

[0113] Step S601, obtaining the user requested target and outputting it according to the user requested target;

[0114] Step S602, correcting the process of outputting the target according to the user's request through the user's login environment data and the login environment recognition model;

[0115] Step S603, obtaining the number of times the user login environment security does not meet the standard within a preset identification period, comparing the number of times the user does not meet the standard with the preset number of times the user does not meet the standard, judging the user's frequency of not meeting the standard based on the comparison result, and outputting the judgment result.

[0116] Specifically, in step S601, the user request target is obtained and outputted according to the user request target, wherein:

[0117] When the user's request target is to authenticate a work of art, a work of art attribute data collection window is pushed to the user, and the user enters the work of art attribute data in the work of art attribute data collection window, and steps S2 to S5 are executed according to the data in the work of art attribute data collection window;

[0118] When the user's request target is to query artworks, obtain the artwork identification code input by the user, and provide the user with a digital authentication certificate corresponding to the artwork identification code input by the user;

[0119] When the user's request target is to trade artworks, the artwork identification code entered by the user is obtained, and output is performed according to the artwork identification code entered by the user, where:

[0120] If the artwork identification code input by the user exists, the artwork attribute data corresponding to the artwork identification code input by the user is provided to the user, and a transaction warning window is pushed to the transaction administrator, who monitors the process of the artwork transaction and obtains the artwork attribute data after the transaction, and executes steps S2 to S5 according to the artwork attribute data after the transaction;

[0121] If the artwork identification code entered by the user does not exist, blank data is provided to the user.

[0122] Specifically, this embodiment does not limit the method for obtaining the user request target. Relevant technical personnel in the field can freely set it according to actual needs, and it is only necessary to meet the need to obtain the user request target. For example, the method for obtaining the user request target can be set to obtain the user request target through the HTTP request header. The artwork attribute data collection window refers to a user interface for collecting, entering and updating relevant attribute information of artworks. The transaction administrator refers to a professional who is responsible for supervising, managing and executing transaction processes on the artwork trading platform. The transaction warning window refers to an interface for issuing warnings to the transaction administrator during the artwork transaction process. This embodiment does not limit the process of the transaction administrator monitoring the process of trading artworks. Relevant technical personnel in the field can freely set it according to actual needs, and it is only necessary to meet the need to monitor the process of trading artworks by the transaction administrator. For example, the method for the transaction administrator to monitor the process of trading artworks can be set to real-time monitoring of transaction data. The blank data refers to disguised meaningless information.

[0123] Specifically, in step S601, an intuitive and easy-to-use user interface and interaction method are provided, so that users can easily complete operations such as artwork authentication, transactions and inquiries. Through automated and digital management processes, the system improves the efficiency of artwork authentication, transactions and inquiries, reduces human errors and costs, and at the same time, transaction monitoring and artwork identification code verification mechanisms ensure the security and reliability of the transaction process, reducing the risk of fraud and illegal transactions.

[0124] Specifically, in step S602, the user login environment data is acquired, the user login environment data is input into the login environment recognition model for recognition, the recognition result is output, the security of the user login environment is judged according to the recognition result, and the process of outputting according to the user request target is corrected according to the judgment result, wherein:

[0125] When the recognition result of the login environment recognition model is that there is no potential safety hazard, it is determined that the security of the user login environment meets the standard, and the process of outputting according to the user request target is not modified;

[0126] When the recognition result of the login environment recognition model is that there is a security risk, it is determined that the security of the user login environment does not meet the standard, and the process of outputting according to the user's request target is corrected to output blank data.

[0127] Specifically, this embodiment does not limit the method for obtaining the user login environment data. Relevant technical personnel in this field can freely set it according to actual needs, and only need to meet the needs of obtaining the user login environment data. For example, the method for obtaining the user login environment data can be set to resolve the user's login environment information through the IP address. The user login environment data refers to the information about the user's login environment collected by the system when the user tries to log in to the system, such as IP address, device information, timestamp, geographic location information and login history. The login environment recognition model refers to a decision tree model that is trained according to a preset environment database and meets a preset recognition accuracy rate. The preset environment database refers to a database that is pre-set and used to train the decision tree model. The decision tree model refers to a model used to classify The present embodiment does not limit the training method of the decision tree model, and relevant technical personnel in the field can freely set it according to actual needs, and only needs to meet the needs of identifying the user login environment data. For example, the training method of the decision tree model can be set to divide the preset environment database into 80% recognition training set and 20% recognition verification set, and input the recognition training set into the decision tree model for training, and input the recognition verification set into the trained decision tree model to iteratively optimize the decision tree model until the recognition result of the decision tree model meets the preset recognition accuracy rate, and then output the decision tree model as a login environment recognition model. The recognition training set refers to a data set used to train the decision tree model, and the recognition verification set refers to a data set used to evaluate and adjust the decision tree model.

[0128] Specifically, in step S602, by introducing the login environment recognition model, potential security threats can be identified more accurately and corresponding measures can be taken to prevent them, thereby significantly improving the overall security. When security risks are found in the user login environment, the output of sensitive information can be automatically stopped, effectively protecting the user's personal privacy and rights.

[0129] Specifically, in step S603, the number of times P of the user login environment security not meeting the standard within the preset identification period is obtained, the number of times P not meeting the standard is compared with the preset number of times P0 not meeting the standard, the frequency of user non-compliance is judged according to the comparison result, and the correction process is optimized according to the judgment result, wherein:

[0130] When P≤P, the user is judged as a low-frequency user that does not meet the standard, and the correction process is not optimized;

[0131] When P>P0, the user is determined to be a high-frequency non-compliant user, the user is marked as a dangerous user, and the recognition result of the login environment recognition model is optimized to have a security risk.

[0132] Specifically, the preset identification period refers to a preset time period used to count and evaluate the number of times the user login environment security does not meet the standards. The present embodiment does not limit the specific value of the preset identification period. Relevant technical personnel in the field can freely set it according to actual needs, and it is only necessary to meet the need to judge the frequency of user non-compliance. For example, the preset value identification period can be set to 7 days. The number of times the user login environment security does not meet the standards refers to the number of times the user login environment is judged by the identification model as having security risks within the preset identification period. The preset number of non-compliance refers to the preset number of non-compliance used to judge whether the number of times the user login environment security does not meet the standards within the preset identification period is too many. The present embodiment does not limit the value of the preset number of non-compliance. Relevant technical personnel in the field can freely set it according to actual needs, and it is only necessary to meet the need to judge the frequency of user non-compliance. For example, the preset number of non-compliance P0 can be set to 3 times.

[0133] Specifically, in step S603, by identifying and strengthening monitoring of high-frequency non-compliant users, potential security threats can be more effectively prevented and overall security can be improved. At the same time, through dynamic risk assessment and personalized security policies, the system can more accurately judge the security of the user's login environment and reduce misjudgments and missed judgments. For low-frequency non-compliant users, the system maintains regular monitoring and does not affect their normal usage experience.

[0134] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. An AID authentication and anti-counterfeiting method based on digital blockchain, characterized in that: include: Step S1, collecting artwork attribute data; Step S2, preprocessing the image data according to a preprocessing method to obtain preprocessed image data; Step S3, performing feature processing on the preprocessed image data to obtain feature image data; Step S4, performing hash processing on the feature image data and non-image data to generate a hash value, and storing the hash value and artwork attribute data on the blockchain to generate a digital authentication certificate; Step S5, constructing an identification code conversion model, and converting the hash value into an artwork identification code through the identification code conversion model; Step S6: output the user operation process according to the security processing method.

2. The AID authentication and anti-counterfeiting method based on digital blockchain according to claim 1 is characterized in that: The pretreatment method comprises: Step S200, resizing and grayscale processing the image data to obtain standard image data; Step S201, performing denoising processing on the standard image data by Gaussian filtering to obtain denoised image data; Step S202: performing edge processing on the denoised image data to obtain pre-processed image data.

3. The AID authentication and anti-counterfeiting method based on digital blockchain according to claim 2 is characterized in that: In the step S200, when the image data is resized and grayscale processed, the image data is resized to a preset size through an image editing tool, and the image data is converted into a grayscale image, and the converted image data is output as standard image data; In step S201, when the standard image data is subjected to denoising processing by Gaussian filtering to obtain denoised image data, the Gaussian filtering calculation formula is set as: Where G(x,y) is the value of the Gaussian function at the point (x,y); x is the distance in the horizontal direction; y is the distance in the vertical direction; π is the ratio of a circle to a circle; e is the base of natural logarithm; σ is the standard deviation of the Gaussian function, and its calculation formula is: Where W is the width of the image; H is the height of the image; In the step S202, when edge processing is performed on the denoised image data, edge processing is performed on the denoised image data by using a Laplace operator to obtain pre-processed image data.

4. The AID authentication and anti-counterfeiting method based on digital blockchain according to claim 1 is characterized in that: In step S3, when feature processing is performed on the preprocessed image data, feature processing is performed on the preprocessed image data by Hough transform, each edge point in the image is mapped to a straight line in the Hough space by using the Hough transform, an intersection point is found in the Hough space, the accumulation degree Q of the intersection point is compared with the preset accumulation degree Q0, the existence of the straight line is judged according to the comparison result, and output is performed according to the judgment result, wherein: When Q<Q0, it is determined that the existence of the straight line does not meet the standard and the straight line is not output; When Q≥Q0, the existence of the straight line is determined to be satisfactory, and the parameters of the straight line are output as feature image data.

5. The AID authentication and anti-counterfeiting method based on digital blockchain according to claim 1 is characterized in that: In step S4, when hashing the feature image data and non-image data to generate a hash value, and storing the hash value and artwork attribute data on the blockchain to form an electronic certificate, the feature image data and non-image data are hashed using SHA-256 in the hash function to generate a hash value, and the hash value, feature image data and non-image data are stored in a preset smart contract, and a digital authentication certificate containing the feature image data, non-image data, hash value, transaction ID and contract address on the blockchain is generated.

6. The AID authentication and anti-counterfeiting method based on digital blockchain according to claim 1 is characterized in that: In step S5, when constructing the identification code conversion model, the artwork identification code conversion database is divided into a conversion training set of 70%, a conversion verification set of 20%, and a conversion test set of 10%. The convolutional neural network model is trained by the conversion verification set, the conversion verification set is input into the trained convolutional neural network model for parameter optimization, and the conversion test set is input into the convolutional neural network model after parameter optimization for testing to obtain the conversion accuracy E, the conversion accuracy E is compared with the preset conversion accuracy E0, the training compliance of the convolutional neural network model is judged according to the comparison result, and the judgment result is output, wherein: When E≥E0, the convolutional neural network model is judged to have met the training standards, the convolutional neural network model is output as an identification code conversion model, and the hash value is converted into an artwork identification code; When E<E0, it is determined that the training of the convolutional neural network model does not meet the standards, and a reminder to update the artwork identification code conversion database is pushed to the administrator, who updates the artwork identification code conversion database and repeats step S5 with the updated artwork identification code conversion database until the training of the convolutional neural network model meets the standards.

7. The AID authentication and anti-counterfeiting method based on digital blockchain according to claim 1 is characterized in that: The safe processing method includes: Step S601, obtaining the user requested target and outputting it according to the user requested target; Step S602, correcting the process of outputting the target according to the user's request through the user's login environment data and the login environment recognition model; Step S603, obtaining the number of times the user login environment security does not meet the standard within a preset identification period, comparing the number of times the user does not meet the standard with the preset number of times the user does not meet the standard, judging the user's frequency of not meeting the standard based on the comparison result, and outputting the judgment result.

8. The AID authentication and anti-counterfeiting method based on digital blockchain according to claim 7 is characterized in that: In step S601, the user request target is obtained and outputted according to the user request target, wherein: When the user's request target is to authenticate a work of art, a work of art attribute data collection window is pushed to the user, and the user enters the work of art attribute data in the work of art attribute data collection window, and steps S2 to S5 are executed according to the data in the work of art attribute data collection window; When the user's request target is to query artworks, obtain the artwork identification code input by the user, and provide the user with a digital authentication certificate corresponding to the artwork identification code input by the user; When the user's request target is to trade artworks, the artwork identification code entered by the user is obtained, and output is performed according to the artwork identification code entered by the user, where: If the artwork identification code input by the user exists, the artwork attribute data corresponding to the artwork identification code input by the user is provided to the user, and a transaction warning window is pushed to the transaction administrator, who monitors the process of the artwork transaction and obtains the artwork attribute data after the transaction, and executes steps S2 to S5 according to the artwork attribute data after the transaction; If the artwork identification code entered by the user does not exist, blank data is provided to the user.

9. The AID authentication and anti-counterfeiting method based on digital blockchain according to claim 7 is characterized in that: In the step S602, the user login environment data is obtained, the user login environment data is input into the login environment recognition model for recognition, the recognition result is output, the security of the user login environment is judged according to the recognition result, and the process of outputting according to the user request target is corrected according to the judgment result, wherein: When the recognition result of the login environment recognition model is that there is no potential safety hazard, it is determined that the security of the user login environment meets the standard, and the process of outputting according to the user request target is not modified; When the recognition result of the login environment recognition model is that there is a security risk, it is determined that the security of the user login environment does not meet the standard, and the process of outputting according to the user's request target is corrected to output blank data.

10. The AID authentication and anti-counterfeiting method based on digital blockchain according to claim 7 is characterized in that: In step S603, the number of times P of the user login environment security not meeting the standard within the preset identification period is obtained, the number of times P not meeting the standard is compared with the preset number of times P0 not meeting the standard, the frequency of user non-compliance is judged according to the comparison result, and the correction process is optimized according to the judgment result, wherein: When P≤P, the user is judged as a low-frequency user that does not meet the standard, and the correction process is not optimized; When P>P0, the user is determined to be a high-frequency non-compliant user, the user is marked as a dangerous user, and the recognition result of the login environment recognition model is optimized to have a security risk.

Images