Multi-level network identity verification and access control system

Abstract

The invention discloses a multi-level network identity verification and access control system, which is characterized in that a biological feature recognition technology is further introduced after traditional password verification, and multi-scale feature extraction is carried out on a fingerprint image input by a user by utilizing an image processing technology based on deep learning; shallow detail features and global structure features of fingerprints are captured, and comprehensive feature representation of a fingerprint image is obtained through feature selection enhancement and multi-level feature joint perception processing, so that a defogged fingerprint image is generated, and based on pixel-by-pixel fingerprint difference information between the defogged fingerprint image and an authorized fingerprint image, the defogged fingerprint image is obtained. A final verification result is generated, and it is ensured that only the verified user can access the system or data conforming to the role of the user. Through the mode, the problem of poor fingerprint image quality can be effectively solved, so that the accuracy and safety of identity verification are improved, and the overall performance of the system and the user experience are optimized.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and more specifically, to a multi-level network identity authentication and access control system. Background Art

[0002] With the rapid development of information technology, network security issues have become increasingly prominent. In a network environment, ensuring the authenticity and legitimacy of user identities is the basis for ensuring system security. The traditional single password authentication method can no longer meet the needs of modern information security, because passwords are easily forgotten, leaked or cracked. In order to improve the security of identity authentication, multi-factor authentication has gradually become a mainstream trend, among which the multi-level identity authentication system that combines password authentication and biometric recognition technology is favored for its higher security and reliability.

[0003] Biometric identification technologies, such as fingerprint recognition, facial recognition, and iris recognition, play an important role in identity authentication due to their uniqueness and resistance to forgery. Among them, fingerprint recognition is one of the most mature and widely used biometric identification technologies. Traditional fingerprint recognition systems usually rely on fingerprint sensors to directly obtain fingerprint images, and then compare them with fingerprint templates in the database through matching algorithms to verify the user's identity. However, existing verification systems usually face problems such as low fingerprint image quality and sensitivity to environmental changes in practical applications. For example, when the finger is wet or stained, the quality of the fingerprint image may be affected, resulting in a decrease in recognition accuracy.

[0004] Therefore, an optimized multi-level network identity authentication and access control system is expected. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a multi-level network identity authentication and access control system, which further introduces biometric recognition technology after traditional password verification, and uses deep learning-based image processing technology to perform multi-scale feature extraction on the fingerprint image input by the user, captures the shallow detail features and global structural features of the fingerprint, and obtains a comprehensive feature representation of the fingerprint image through feature selection enhancement and multi-level feature joint perception processing, thereby generating a defogging fingerprint image, and then based on the pixel-by-pixel fingerprint difference information between the defogging fingerprint image and the authorized fingerprint image, generates the final verification result, ensuring that only verified users can access systems or data that match their roles. In this way, the problem of poor fingerprint image quality can be effectively addressed, thereby improving the accuracy and security of identity authentication and optimizing the overall performance and user experience of the system.

[0006] According to one aspect of the present application, a multi-level network identity authentication and access control system is provided, comprising:

[0007] User account information acquisition module, used to obtain the user name and password entered by the user;

[0008] A user account verification module, used to perform a first level of verification based on the user name and the password;

[0009] A user fingerprint receiving module, used to receive a fingerprint image input by a user after the first layer verification is passed;

[0010] A user fingerprint verification module, used to perform a second layer of verification based on the fingerprint image;

[0011] An access request receiving module, used for receiving an access request input by the user after the second layer verification is passed;

[0012] An access rights verification module, used to verify the user's access rights according to the RBAC policy;

[0013] Wherein, the user fingerprint verification module includes:

[0014] A fingerprint image multi-scale feature extraction unit, used for performing multi-scale feature extraction on the fingerprint image to obtain a fingerprint shallow feature map and a fingerprint structure feature map;

[0015] A feature selection and strengthening unit, used for performing feature selection and strengthening processing on the fingerprint shallow feature map and the fingerprint structure feature map respectively to obtain a fingerprint shallow layer enhanced feature map and a fingerprint structure enhanced feature map;

[0016] A defogging fingerprint image generating unit, configured to generate a defogging fingerprint image based on the joint features of the fingerprint shallow layer enhancement feature map and the fingerprint structure enhancement feature map;

[0017] The verification result generating unit is used to determine whether the fingerprint authentication is passed based on the pixel difference between the defogging fingerprint image and the authorized fingerprint image.

[0018] Compared with the prior art, the multi-level network identity authentication and access control system provided by the present application further introduces biometric recognition technology after traditional password verification, and uses deep learning-based image processing technology to perform multi-scale feature extraction on the fingerprint image input by the user, capturing the shallow detail features and global structural features of the fingerprint, and through feature selection enhancement and multi-level feature joint perception processing, obtains a comprehensive feature representation of the fingerprint image, thereby generating a defogging fingerprint image, and then based on the pixel-by-pixel fingerprint difference information between the defogging fingerprint image and the authorized fingerprint image, generates the final verification result, ensuring that only verified users can access the system or data that matches their role. In this way, the problem of poor fingerprint image quality can be effectively addressed, thereby improving the accuracy and security of identity authentication, and optimizing the overall performance and user experience of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0020] Figure 1 A block diagram of a multi-level network identity authentication and access control system according to an embodiment of the present application;

[0021] Figure 2 A schematic diagram of data flow of a multi-level network identity authentication and access control system according to an embodiment of the present application;

[0022] Figure 3 It is a block diagram of a user fingerprint verification module in a multi-level network identity authentication and access control system according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0024] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0025] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0026] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0027] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0028] The traditional single password authentication method can no longer meet the needs of modern information security, because passwords are easily forgotten, leaked or cracked. In order to improve the security of identity authentication, multi-factor authentication has gradually become a mainstream trend. Among them, the multi-level identity authentication system that combines password authentication and biometric recognition technology is favored for its higher security and reliability. Biometric recognition technologies, such as fingerprint recognition, facial recognition, iris recognition, etc., play an important role in identity authentication due to their uniqueness and difficulty in forging. Among them, fingerprint recognition is one of the most mature and widely used biometric recognition technologies. Traditional fingerprint recognition systems usually rely on fingerprint sensors to directly obtain fingerprint images, and then compare them with fingerprint templates in the database through matching algorithms to verify the identity of the user. However, existing verification systems usually face problems such as low fingerprint image quality and sensitivity to environmental changes in practical applications. For example, when the finger is wet or stained, the quality of the fingerprint image may be affected, resulting in a decrease in recognition accuracy. Therefore, an optimized multi-level network identity authentication and access control system is expected.

[0029] In the technical solution of the present application, a multi-level network identity authentication and access control system is proposed. Figure 1 It is a block diagram of a multi-level network identity authentication and access control system according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of a multi-level network identity authentication and access control system according to an embodiment of the present application. Figure 1 and Figure 2As shown, according to the multi-level network identity authentication and access control system 300 of the embodiment of the present application, it includes: a user account information acquisition module 310, used to obtain a user name and password input by a user; a user account verification module 320, used to perform a first-level verification based on the user name and the password; a user fingerprint receiving module 330, used to receive a fingerprint image input by the user after the first-level verification is passed; a user fingerprint verification module 340, used to perform a second-level verification based on the fingerprint image; an access request receiving module 350, used to receive an access request input by the user after the second-level verification is passed; an access permission verification module 360, used to verify the user's access permission according to the RBAC policy.

[0030] In particular, the user account information acquisition module 310 and the user account verification module 320 are used to obtain the user name and password input by the user; based on the user name and the password, a first level of verification is performed. By verifying the user name and password, the system can ensure that only authorized users can access protected resources or systems.

[0031] In particular, the user fingerprint receiving module 330 is used to receive a fingerprint image input by a user after the first layer of verification is passed. It should be understood that fingerprints are highly unique biometric features that can help organizations ensure that only authorized users can access protected resources or systems. After the first layer of verification (usually involving a username and password) is passed, receiving a fingerprint image input by a user can further enhance security.

[0032] In particular, the user fingerprint verification module 340 is used to perform a second layer of verification based on the fingerprint image. In particular, in a specific example of the present application, Figure 3 As shown, the user fingerprint verification module 340 includes: a fingerprint image multi-scale feature extraction unit 341, which is used to perform multi-scale feature extraction on the fingerprint image to obtain a fingerprint shallow feature map and a fingerprint structure feature map; a feature selection and enhancement unit 342, which is used to perform feature selection and enhancement processing on the fingerprint shallow feature map and the fingerprint structure feature map respectively to obtain a fingerprint shallow enhancement feature map and a fingerprint structure enhancement feature map; a defogging fingerprint image generation unit 343, which is used to generate a defogging fingerprint image based on the joint features of the fingerprint shallow enhancement feature map and the fingerprint structure enhancement feature map; a verification result generation unit 344, which is used to determine whether the fingerprint authentication is passed based on the pixel difference between the defogging fingerprint image and the authorized fingerprint image.

[0033] Specifically, the fingerprint image multi-scale feature extraction unit 341 is used to perform multi-scale feature extraction on the fingerprint image to obtain a fingerprint shallow feature map and a fingerprint structural feature map. In a specific example of the present application, the fingerprint image is input into a multi-scale fingerprint feature scanner based on a dilated pyramid network to obtain the fingerprint shallow feature map and the fingerprint structural feature map. Considering that the fingerprint image contains rich feature information, there are both shallow features with rich details (such as local patterns, breakpoints, bifurcation points, etc. of ridges and valleys) and deep features reflecting the global structure of the fingerprint (such as overall texture patterns, direction fields, etc.). Therefore, in order to fully capture the multi-level features of the fingerprint image, the present application uses a multi-scale fingerprint feature scanner based on a dilated pyramid network to process the fingerprint image, and uses dilated convolution kernels of different scales to extract feature information of different levels of the fingerprint image layer by layer to capture the shallow detail features and global structural features in the fingerprint image, and generate a fingerprint shallow feature map and a fingerprint structural feature map. By comprehensively utilizing the multi-scale feature information of the fingerprint image, the adaptability of the model to fingerprint images of different qualities can be effectively enhanced, thereby more accurately performing defogging and reconstruction of the fingerprint image.

[0034] Specifically, the feature selection and enhancement unit 342 is used to perform feature selection and enhancement processing on the fingerprint shallow feature map and the fingerprint structure feature map respectively to obtain the fingerprint shallow enhancement feature map and the fingerprint structure enhancement feature map. In a specific example of the present application, the fingerprint shallow feature map and the fingerprint structure feature map are input into a feature enhancement module based on the characteristic energy continuous distribution characteristics to obtain the fingerprint shallow enhancement feature map and the fingerprint structure enhancement feature map. Considering that in the actual acquisition process of the fingerprint image, there may be various noise interferences, such as the shot noise introduced by the fingerprint acquisition device, the low-frequency noise caused by uneven illumination, and the additional texture noise in the image caused by the different finger placement positions, angles and pressures of the user when collecting fingerprints. Therefore, in order to further improve the accuracy of fingerprint image defogging and reconstruction, the present application introduces a feature enhancement module based on the characteristic energy continuous distribution characteristics to perform feature selection and enhancement processing on the fingerprint shallow feature map and the fingerprint structure feature map respectively.

[0035] In an embodiment of the present application, the fingerprint shallow feature map and the fingerprint structure feature map are input into a feature enhancement module based on the characteristic energy continuous distribution characteristics to obtain the fingerprint shallow enhancement feature map and the fingerprint structure enhancement feature map, including: first, splitting the fingerprint shallow feature map into a set of fingerprint shallow local feature vectors; then, calculating the characteristic energy distribution spectrum vector of each fingerprint shallow local feature vector in the set of fingerprint shallow local feature vectors to obtain a set of fingerprint shallow feature energy distribution spectrum vectors; by calculating the characteristic energy distribution spectrum vector of each local feature vector, the distribution of each local feature vector at different energy levels is revealed; then, the positional mean vector of the set of fingerprint shallow feature energy distribution spectrum vectors is calculated as the energy spectrum distribution center vector; here, the positional mean of the characteristic energy distribution spectrum vector of each local feature vector is calculated to reveal the overall energy distribution trend of the feature map and obtain the energy spectrum distribution center vector; then, the energy The Mahalanobis distance between the spectral distribution center vector and each fingerprint shallow feature energy distribution spectrum vector in the set of fingerprint shallow feature energy distribution spectrum vectors is used as the energy distribution spectrum span factor to obtain a set of energy distribution spectrum span factors; that is, by calculating the difference between the energy spectrum distribution center vector and each feature energy distribution spectrum vector, the closeness of each local feature vector to the overall distribution is evaluated, and the corresponding energy distribution spectrum span factor is generated; further, based on the comparison between each energy distribution spectrum span factor in the set of energy distribution spectrum span factors and a preset threshold, the fingerprint shallow local feature vectors corresponding to the energy distribution spectrum span factor less than the preset threshold are selected to form a set of selected fingerprint shallow local feature vectors; in this way, redundant information deviating from the overall distribution is removed, and useful features close to the overall distribution are retained; then the selected set of fingerprint shallow local feature vectors is trans-aggregated in a feature splitting manner to obtain the fingerprint shallow enhanced feature map.

[0036] Among them, calculating the characteristic energy distribution spectrum vector of each fingerprint shallow local feature vector in the set of fingerprint shallow local feature vectors to obtain a set of fingerprint shallow feature energy distribution spectrum vectors, including: calculating the characteristic distribution energy synergy representation vector between the i-th fingerprint shallow local feature vector and each other fingerprint shallow local feature vector in the set of fingerprint shallow local feature vectors in a manner of taking the maximum eigenvalue element by element to obtain a set of feature distribution energy synergy representation vectors; calculating the characteristic distribution energy synergy factor of each feature distribution energy synergy representation vector in the set of feature distribution energy synergy representation vectors to obtain the characteristic energy distribution spectrum vector of the i-th fingerprint shallow local feature vector composed of multiple feature distribution energy synergy factors. More specifically, the process of calculating the feature distribution energy synergy factor of each feature distribution energy synergy representation vector in the set of feature distribution energy synergy representation vectors to obtain the feature energy distribution spectrum vector of the i-th fingerprint shallow local feature vector composed of multiple feature distribution energy synergy factors includes: calculating the square of the difference between the maximum eigenvalue of the feature distribution energy synergy representation vector and its feature mean, the feature variance of the feature distribution energy synergy representation vector, and the weighted sum of the bias terms, and then adding the variance of the feature distribution energy synergy representation vector and the bias term and dividing the sum by the weighted sum to obtain the feature distribution energy synergy factor.

[0037] In summary, in the above embodiment, the fingerprint shallow feature map and the fingerprint structure feature map are respectively subjected to feature selection and enhancement processing to obtain a fingerprint shallow enhanced feature map and a fingerprint structure enhanced feature map, including: processing the set of fingerprint shallow local feature vectors with the following feature selection optimization formula to obtain the set of selected fingerprint shallow local feature vectors, wherein the feature selection optimization formula is:

[0038] X i,j =max(X i ,X j )

[0039]

[0040] P i ={e i,1 ,e i,2 ,...,e i,m}

[0041]

[0042]

[0043] X′={V i |Dist(c,P i )≤θ}

[0044] Among them, X i and X j denote the ith and jth shallow local feature vectors of the fingerprint in the set of shallow local feature vectors of the fingerprint, respectively; max(·,·) denotes taking the maximum eigenvalue of the two feature vectors element by element; X i,j The feature distribution energy synergy representation vector between the i-th fingerprint shallow local feature vector and the j-th fingerprint shallow local feature vector, μ, σ 2 and max(V i,j ) represent the characteristic mean, characteristic variance and maximum eigenvalue of the characteristic distribution energy cooperative representation vector, ε is the bias term, e i,1 、e i,2 、e i,j and e i,m Respectively represent the characteristic distribution energy synergy factors between the i-th fingerprint shallow local feature vector and the first, second, j-th and m-th fingerprint shallow local feature vectors in the set of the fingerprint shallow local feature vectors, where the value of m is the number of the fingerprint shallow local feature vectors, P i represents the energy distribution spectrum vector of the shallow feature of the i-th fingerprint, P represents the set of energy distribution spectrum vectors of the shallow feature of the fingerprint, c represents the center vector of the energy spectrum distribution, S i,c represents the covariance matrix between the energy distribution spectrum vector of the shallow feature of the i-th fingerprint and the center vector of the energy spectrum distribution, Dist(·,·) represents the distance metric function, D i represents the span factor of the ith energy distribution spectrum, θ represents the preset threshold, and X′ represents the set of shallow local feature vectors of the selected fingerprint.

[0045] Specifically, the defogging fingerprint image generation unit 343 is used to generate a defogging fingerprint image based on the joint features of the fingerprint shallow enhancement feature map and the fingerprint structure enhancement feature map. In a specific example of the present application, the fingerprint shallow enhancement feature map and the fingerprint structure enhancement feature map are first input into a feature joint saliency perception network based on attention-driven to obtain a fingerprint shallow-structure saliency joint perception feature map; and then the fingerprint shallow-structure saliency joint perception feature map is input into a fingerprint defogging device based on a diffusion model to obtain the defogging fingerprint image. It should be understood that the fingerprint shallow enhancement feature map mainly describes the basic texture and edge details of the fingerprint image, while the fingerprint structure enhancement feature map focuses more on describing the overall structure and directional field information of the fingerprint image. In order to combine the advantages of both and further improve the defogging effect of the fingerprint image, the present application introduces a feature joint saliency perception network based on attention-driven to perform feature joint processing on the fingerprint shallow enhancement feature map and the fingerprint structure enhancement feature map to form a more comprehensive and rich fingerprint feature representation. Specifically, the network first reshapes the fingerprint shallow enhancement feature map and the fingerprint structural enhancement feature map into feature matrices so that their dimensions and shapes are adapted to the input requirements of the attention mechanism. Then, by calculating the bidirectional dependency between the two sets of feature matrices, the interaction between the features is quantified, and the fingerprint structure-shallow feature and fingerprint shallow-structural feature dependency matrices are generated, which serve as the guiding basis for feature attention optimization fusion. At the same time, in order to reduce the model's sensitivity to noise and prevent overfitting, the two sets of dependency matrices are further pruned to enhance the generalization ability of the model by selectively ignoring or suppressing part of the feature interaction. Furthermore, based on the pruned two sets of dependency matrices, the two sets of feature matrices after shape reshaping are weighted optimized, and then restored to the original feature map form by shape reshaping and weighted fusion to generate the fingerprint shallow-structural significant joint perception feature map. Through this dual attention interaction mechanism, the correlation dependency between the shallow features and structural features of the fingerprint image can be effectively mined, and the in-depth fusion of multi-scale features can be achieved, thereby further improving the defogging effect of the fingerprint image. Furthermore, the fingerprint shallow-structure significant joint perception feature map is input into a fingerprint defogging device based on a diffusion model for image defogging and reconstruction. A person skilled in the art should know that a diffusion model is a mathematical model based on a physical diffusion process, which can generate images by simulating a diffusion process from a high entropy state to a low entropy state. In the technical solution of the present application, the diffusion model is based on the fingerprint feature information in the fingerprint shallow-structure significant joint perception feature map, and gradually adjusts the pixel values ​​in the feature map through an iterative optimization algorithm to remove noise and blur, restore image details, and retain and enhance the key information in the original feature map, thereby generating a clear defogging fingerprint image.

[0046] The method comprises: inputting the fingerprint shallow enhancement feature map and the fingerprint structural enhancement feature map into a feature joint saliency perception network based on attention driving to obtain a fingerprint shallow-structural saliency joint perception feature map, including: inputting the fingerprint shallow enhancement feature map and the fingerprint structural enhancement feature map into a feature joint saliency perception network based on attention driving to obtain a fingerprint shallow-structural saliency joint perception feature map; and inputting the fingerprint shallow-structural saliency joint perception feature map into a fingerprint dehazer based on a diffusion model to obtain the dehazed fingerprint image. The process of inputting the fingerprint shallow enhancement feature map and the fingerprint structure enhancement feature map into the attention-driven feature joint saliency perception network to obtain the fingerprint shallow-structure saliency joint perception feature map includes: reshaping the fingerprint shallow enhancement feature map and the fingerprint structure enhancement feature map to obtain the fingerprint shallow enhancement feature shape reshaping matrix and the fingerprint structure enhancement feature shape reshaping matrix; inputting the fingerprint shallow enhancement feature shape reshaping matrix and the fingerprint structure enhancement feature shape reshaping matrix into the feature channel-by-channel interactive perception module to obtain the fingerprint shallow-structure feature dependency matrix and the fingerprint structure-shallow feature dependency matrix; inputting the fingerprint shallow-structure feature dependency matrix and the fingerprint structure-shallow feature dependency matrix into the random inactivation module to obtain the pruned fingerprint shallow-structure feature dependency matrix. The invention relates to a method for obtaining a dependency relationship optimized fingerprint structure enhancement feature matrix by matrix multiplying the pruned fingerprint shallow-structure feature dependency matrix with the fingerprint structure enhancement feature shape reshaping matrix to obtain a dependency relationship optimized fingerprint shallow enhancement feature matrix; the pruned fingerprint structure-shallow feature dependency matrix is ​​matrix multiplied with the fingerprint shallow enhancement feature shape reshaping matrix to obtain a dependency relationship optimized fingerprint shallow enhancement feature matrix; the dependency optimized fingerprint shallow enhancement feature matrix and the dependency optimized fingerprint structure enhancement feature matrix are feature reshaped to obtain an optimized fingerprint shallow enhancement feature map and an optimized fingerprint structure enhancement feature map; the weighted sum of the optimized fingerprint shallow enhancement feature map and the optimized fingerprint structure enhancement feature map is calculated to obtain the fingerprint shallow-structure significant joint perception feature map.More specifically, the process of inputting the fingerprint shallow enhancement feature shape reshaping matrix and the fingerprint structural enhancement feature shape reshaping matrix into the feature channel-by-channel interactive perception module to obtain the fingerprint shallow-structural feature dependency matrix and the fingerprint structure-shallow feature dependency matrix includes: calculating the fingerprint shallow enhancement feature shape reshaping matrix multiplied by the transposed matrix of the fingerprint structure enhancement feature shape reshaping matrix to obtain the fingerprint shallow-structural feature association representation matrix; dividing the fingerprint shallow-structural feature association representation matrix by the scale of the fingerprint structure enhancement feature shape reshaping matrix and then inputting it into the softmax function to obtain the fingerprint shallow-structural feature dependency matrix; calculating the transposed matrix of the fingerprint shallow enhancement feature shape reshaping matrix multiplied by the fingerprint structure enhancement feature shape reshaping matrix to obtain the fingerprint structure-shallow feature association representation matrix; dividing the fingerprint structure-shallow feature association representation matrix by the scale of the fingerprint shallow enhancement feature shape reshaping matrix and then inputting it into the softmax function to obtain the fingerprint structure-shallow feature dependency matrix.

[0047] More specifically, the fingerprint shallow layer enhancement feature map and the fingerprint structure enhancement feature map are input into the attention-driven feature joint saliency perception network to obtain the fingerprint shallow layer-structure saliency joint perception feature map, including: processing the fingerprint shallow layer enhancement feature map and the fingerprint structure enhancement feature map with the following joint perception formula to obtain the fingerprint shallow layer-structure saliency joint perception feature map, wherein the joint perception formula is:

[0048] M1=Reshape(F1)

[0049] M2=Reshape(F2)

[0050]

[0051]

[0052] W 21 = Dropout(M 21 )

[0053] W 12 = Dropout(M 12 )

[0054]

[0055]

[0056] F1'=Reshape(M 1′ )

[0057] F2'=Reshape(M 2′ )

[0058] F c =αF1'+βF2'

[0059] Wherein, F1 represents the fingerprint shallow layer enhancement feature map, F2 represents the fingerprint structure enhancement feature map, reshape(·) represents feature shape reshaping, M1 represents the fingerprint shallow layer enhancement feature shape reshaping matrix, M2 represents the fingerprint structure enhancement feature shape reshaping matrix, represents matrix multiplication, (·) T represents the transpose of the matrix, S represents the scale of the fingerprint shallow enhancement feature shape reshaping matrix, that is, the width of the fingerprint shallow enhancement feature shape reshaping matrix multiplied by the height, and the scale of the fingerprint shallow enhancement feature shape reshaping matrix and the fingerprint structural enhancement feature shape reshaping matrix are the same, Softmax is a normalized exponential function, M 21 represents the fingerprint shallow-structural feature dependency matrix, M 12 represents the fingerprint structure-shallow feature dependency matrix, Dropout(·) represents random inactivation processing, W 21 represents the pruned fingerprint shallow-structural feature dependency matrix, W 12 represents the pruned fingerprint structure-shallow feature dependency matrix, M 1′ Represents the dependency optimization fingerprint shallow enhancement feature matrix, M 2′ represents the dependency relationship optimized fingerprint structure enhanced feature matrix, F1' represents the optimized fingerprint shallow enhanced feature map, F2' represents the optimized fingerprint structure enhanced feature map, α and β represent different weight coefficients respectively, F c Represents the fingerprint shallow-structural significant joint perceptual feature map.

[0060] Specifically, the verification result generating unit 344 is used to determine whether the fingerprint authentication is passed based on the pixel difference between the defogging fingerprint image and the authorized fingerprint image. In a specific example of the present application, the fingerprint pixel-by-pixel difference matrix between the defogging fingerprint image and the authorized fingerprint image is first calculated; then the fingerprint pixel-by-pixel difference matrix is ​​input into the verification result generator based on the classifier to obtain the second layer of verification results, and the second layer of verification results is used to indicate whether the fingerprint authentication is passed. In the technical solution of the present application, the fingerprint pixel-by-pixel difference matrix between the defogging fingerprint image and the authorized fingerprint image is calculated by pixel-by-pixel comparison to reveal the pixel value difference between the defogging fingerprint image and the authorized fingerprint image at the corresponding spatial position, thereby quantitatively describing the similarity between the defogging fingerprint image and the authorized fingerprint image. Then, based on the difference information in the fingerprint pixel-by-pixel difference matrix, the fingerprint pixel-by-pixel difference matrix is ​​input into the verification result generator based on the classifier for classification judgment to obtain the second layer of verification results, and the second layer of verification results is used to indicate whether the fingerprint authentication is passed.

[0061] In a preferred example, inputting the fingerprint shallow-structural significant joint perceptual feature map into a fingerprint defogging device based on a diffusion model to obtain the defogging fingerprint image comprises the following steps:

[0062] Calculate the sum of the absolute values ​​of each eigenvalue of the fingerprint shallow-structure significant joint perception feature map to obtain a first fingerprint shallow-structure significant joint perception spatial structure value, and calculate the square root of the sum of the squares of each eigenvalue of the fingerprint shallow-structure significant joint perception feature map to obtain a second fingerprint shallow-structure significant joint perception spatial structure value;

[0063] Multiply each eigenvalue of the fingerprint shallow-structure significant joint perception feature map by the first fingerprint shallow-structure significant joint perception spatial structure value and the second fingerprint shallow-structure significant joint perception spatial structure value to obtain a first fingerprint shallow-structure significant joint perception structure reference value and a second fingerprint shallow-structure significant joint perception structure reference value corresponding to each eigenvalue;

[0064] Multiplying each eigenvalue of the fingerprint shallow-structure significant joint perceptual feature map by the scale of the fingerprint shallow-structure significant joint perceptual feature map and the square root of the scale to obtain a first fingerprint shallow-structure significant joint perceptual scale transformation value and a second fingerprint shallow-structure significant joint perceptual scale transformation value corresponding to each eigenvalue;

[0065] Dividing the first fingerprint shallow-structure significant joint perception structure reference value by the difference between the first fingerprint shallow-structure significant joint perception space structure value and the first fingerprint shallow-structure significant joint perception scale transformation value to obtain the first fingerprint shallow-structure significant joint perception transformation adjustment value;

[0066] Dividing the second fingerprint shallow-structure significant joint perception structure reference value by the difference between the second fingerprint shallow-structure significant joint perception space structure value and the second fingerprint shallow-structure significant joint perception scale transformation value to obtain the second fingerprint shallow-structure significant joint perception transformation adjustment value;

[0067] Calculating a weighted sum of the first fingerprint shallow-structure significant joint perceptual transformation adjustment value and the second fingerprint shallow-structure significant joint perceptual transformation adjustment value to obtain each feature value of the optimized fingerprint shallow-structure significant joint perceptual feature map;

[0068] The optimized fingerprint shallow-structure significant joint perceptual feature map is input into the fingerprint defogging device based on the diffusion model to obtain the defogging fingerprint image.

[0069] Here, the fingerprint shallow-structure significant joint perceptual feature map, denoted as F, is optimized as follows:

[0070] F′=F1⊕(ω⊙F2)

[0071]

[0072] f 2i =(α×f i ) / (α-n×f i )

[0073]

[0074] n=W×H×C

[0075] Wherein, F is the fingerprint shallow-structure significant joint perceptual feature map, R represents a real number set, W, H and C represent the width, height and number of channels of the matrix respectively, n represents the scale of the fingerprint shallow-structure significant joint perceptual feature map, f irepresents the feature value of the ith position of the fingerprint shallow-structure significant joint perception feature map, α represents the shallow-structure significant joint perception space structure value of the first fingerprint, β represents the shallow-structure significant joint perception space structure value of the second fingerprint, ⊙ represents the point multiplication by position, ⊕ represents the addition by position, and f 1i represents the second fingerprint shallow-structure significant joint perceptual transformation adjustment value corresponding to each eigenvalue, f1 represents the second fingerprint shallow-structure significant joint perceptual transformation adjustment feature map composed of each of the second fingerprint shallow-structure significant joint perceptual transformation adjustment values, f 2i represents the first fingerprint shallow-structural significant joint perceptual transformation adjustment value corresponding to each eigenvalue, F2 represents the first fingerprint shallow-structural significant joint perceptual transformation adjustment feature map composed of the first fingerprint shallow-structural significant joint perceptual transformation adjustment values, ω represents the weighted hyperparameter, and F′ represents the optimized fingerprint shallow-structural significant joint perceptual feature map.

[0076] That is, considering that the fingerprint shallow enhancement feature map and the fingerprint structural enhancement feature map represent the image shallow enhancement semantic features and image deep structural semantic features of the fingerprint image respectively, when further performing attention-driven feature joint saliency perception, the difference in attention weights of the fusion of image shallow features and image deep structural features will cause image semantic saliency joint alignment conflicts, thereby causing the loss of aggregated key information and affecting the expression effect of the fingerprint shallow-structural saliency joint perception feature map.

[0077] Therefore, in the preferred example, for the spatial structure information of the feature set of the fingerprint shallow-structure significant joint perceptual feature map in the high-dimensional space, the scale-based frame transformation of each feature value of the fingerprint shallow-structure significant joint perceptual feature map is performed by taking the class-norm space structured representation of the fingerprint shallow-structure significant joint perceptual feature map as a reference window, and the spatial structure-based frame attention weight adjustment of each feature value of the fingerprint shallow-structure significant joint perceptual feature map is realized to ensure the invariance of the spatial transformation (translation, scaling and rotation) of the fingerprint shallow-structure significant joint perceptual feature map under the interaction of feature space, thereby constructing a solid alignment guardrail of the optimized fingerprint shallow-structure significant joint perceptual feature map based on the aggregated features and the original feature set as a whole to prevent the loss of aggregated key information. In this way, the convergence and generalization effect of the generative regression of the feature set of the fingerprint shallow-structure significant joint perceptual feature map under the complex spatial structure representation is improved, and the image quality of the grayscale image of the aligned assembly state after distortion correction obtained by inputting it into the distortion correction module based on the diffusion model is improved.

[0078] In particular, the access request receiving module 350 and the access permission verification module 360 ​​are used to receive the access request input by the user after the second layer verification is passed; and verify the access permission of the user according to the RBAC policy. In a specific example of the present application, after the second layer verification is passed, the access request input by the user is verified according to the RBAC policy (role-based access control policy) to see if it is consistent with the access permission, thereby ensuring that the user can only access the resources for which he is authorized. For example, financial personnel can only access the financial system, and human resources personnel can only access the human resources system.

[0079] As described above, the multi-level network identity authentication and access control system 300 according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a multi-level network identity authentication and access control algorithm. In a possible implementation, the multi-level network identity authentication and access control system 300 according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the multi-level network identity authentication and access control system 300 can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the multi-level network identity authentication and access control system 300 can also be one of the many hardware modules of the wireless terminal.

[0080] Alternatively, in another example, the multi-level network identity authentication and access control system 300 and the wireless terminal may also be separate devices, and the multi-level network identity authentication and access control system 300 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0081] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A multi-level network identity authentication and access control system, characterized in that: include: User account information acquisition module, used to obtain the user name and password entered by the user; A user account verification module, used to perform a first level of verification based on the user name and the password; A user fingerprint receiving module, used to receive a fingerprint image input by a user after the first layer verification is passed; A user fingerprint verification module, used to perform a second layer of verification based on the fingerprint image; An access request receiving module, used for receiving an access request input by the user after the second layer verification is passed; An access rights verification module, used to verify the user's access rights according to the RBAC policy; Wherein, the user fingerprint verification module includes: A fingerprint image multi-scale feature extraction unit, used for performing multi-scale feature extraction on the fingerprint image to obtain a fingerprint shallow feature map and a fingerprint structure feature map; A feature selection and strengthening unit, used for performing feature selection and strengthening processing on the fingerprint shallow feature map and the fingerprint structure feature map respectively to obtain a fingerprint shallow layer enhanced feature map and a fingerprint structure enhanced feature map; A defogging fingerprint image generating unit, configured to generate a defogging fingerprint image based on the joint features of the fingerprint shallow layer enhancement feature map and the fingerprint structure enhancement feature map; The verification result generating unit is used to determine whether the fingerprint authentication is passed based on the pixel difference between the defogging fingerprint image and the authorized fingerprint image.

2. The multi-level network identity authentication and access control system according to claim 1, characterized in that: The fingerprint image multi-scale feature extraction unit is used to: The fingerprint image is input into a multi-scale fingerprint feature scanner based on a dilated pyramid network to obtain the fingerprint shallow feature map and the fingerprint structural feature map.

3. The multi-level network identity authentication and access control system according to claim 2, characterized in that: The feature selection and strengthening unit is used to: The fingerprint shallow feature map and the fingerprint structure feature map are input into a feature enhancement module based on the characteristic of continuous distribution of feature energy to obtain the fingerprint shallow layer enhancement feature map and the fingerprint structure enhancement feature map.

4. The multi-level network identity authentication and access control system according to claim 3, characterized in that: The feature selection and strengthening unit comprises: A feature splitting subunit, used for splitting the fingerprint shallow feature map into a set of fingerprint shallow local feature vectors; An energy distribution spectrum calculation subunit, used for calculating the characteristic energy distribution spectrum vector of each fingerprint shallow local feature vector in the set of fingerprint shallow local feature vectors to obtain a set of fingerprint shallow feature energy distribution spectrum vectors; An energy spectrum distribution center calculation subunit, used to calculate the position mean vector of the set of energy distribution spectrum vectors of the fingerprint shallow feature as the energy spectrum distribution center vector; An energy distribution spectrum span calculation subunit, used for calculating the Mahalanobis distance between the energy spectrum distribution center vector and each fingerprint shallow feature energy distribution spectrum vector in the set of fingerprint shallow feature energy distribution spectrum vectors as an energy distribution spectrum span factor to obtain a set of energy distribution spectrum span factors; A feature screening subunit, for selecting fingerprint shallow local feature vectors corresponding to energy distribution spectrum span factors less than the preset threshold value based on the comparison between each energy distribution spectrum span factor in the set of energy distribution spectrum span factors and the preset threshold value to form a set of selected fingerprint shallow local feature vectors; The feature aggregation subunit is used to perform reverse aggregation on the selected set of shallow local feature vectors of the fingerprint in a feature splitting manner to obtain the fingerprint shallow enhanced feature map.

5. The multi-level network identity authentication and access control system according to claim 4, characterized in that: The energy distribution spectrum calculation subunit comprises: The feature distribution collaborative representation secondary subunit is used to calculate the feature distribution energy collaborative representation vector between the i-th fingerprint shallow local feature vector and each other fingerprint shallow local feature vector in the set of fingerprint shallow local feature vectors in a manner of taking the maximum eigenvalue element by element to obtain a set of feature distribution energy collaborative representation vectors; The feature distribution energy synergy factor calculation secondary subunit is used to calculate the feature distribution energy synergy factor of each feature distribution energy synergy representation vector in the set of feature distribution energy synergy representation vectors to obtain the feature energy distribution spectrum vector of the i-th fingerprint shallow local feature vector composed of multiple feature distribution energy synergy factors.

6. The multi-level network identity authentication and access control system according to claim 5, characterized in that: The characteristic distribution energy synergy factor calculation secondary subunit is used to: The square of the difference between the maximum eigenvalue of the feature distribution energy synergistic representation vector and its eigenmean, the feature variance of the feature distribution energy synergistic representation vector, and the weighted sum of the bias terms are calculated, and then the variance of the feature distribution energy synergistic representation vector and the bias term are added and divided by the weighted sum to obtain the feature distribution energy synergy factor.

7. The multi-level network identity authentication and access control system according to claim 6, characterized in that: The defogging fingerprint image generating unit comprises: A feature joint perception subunit, used for inputting the fingerprint shallow layer enhanced feature map and the fingerprint structure enhanced feature map into an attention-driven feature joint saliency perception network to obtain a fingerprint shallow layer-structure saliency joint perception feature map; The fingerprint defogging subunit is used to input the fingerprint shallow-structure significant joint perception feature map into a fingerprint defogging device based on a diffusion model to obtain the defogging fingerprint image.

8. The multi-level network identity authentication and access control system according to claim 7, characterized in that: The feature joint perception subunit includes: The first feature shape reshaping secondary subunit is used to perform feature shape reshaping on the fingerprint shallow layer enhancement feature map and the fingerprint structure enhancement feature map to obtain a fingerprint shallow layer enhancement feature shape reshaping matrix and a fingerprint structure enhancement feature shape reshaping matrix; A bidirectional dependency calculation secondary subunit is used to input the fingerprint shallow layer enhancement feature shape reshaping matrix and the fingerprint structure enhancement feature shape reshaping matrix into a feature channel-by-channel interactive perception module to obtain a fingerprint shallow layer-structure feature dependency matrix and a fingerprint structure-shallow feature dependency matrix; A random dropout secondary subunit, used for inputting the fingerprint shallow-structure feature dependency matrix and the fingerprint structure-shallow feature dependency matrix into a random dropout module to obtain a pruned fingerprint shallow-structure feature dependency matrix and a pruned fingerprint structure-shallow feature dependency matrix; A structural feature dependency optimization secondary subunit is used to perform matrix multiplication of the pruned fingerprint shallow-structural feature dependency matrix and the fingerprint structural enhancement feature shape reshaping matrix to obtain a dependency optimization fingerprint structural enhancement feature matrix; A shallow feature dependency optimization secondary subunit is used to perform matrix multiplication of the pruned fingerprint structure-shallow feature dependency matrix and the fingerprint shallow enhancement feature shape reshaping matrix to obtain a dependency optimized fingerprint shallow enhancement feature matrix; The second feature shape reshaping secondary subunit is used to perform feature shape reshaping on the dependency optimized fingerprint shallow layer enhancement feature matrix and the dependency optimized fingerprint structure enhancement feature matrix to obtain an optimized fingerprint shallow layer enhancement feature map and an optimized fingerprint structure enhancement feature map; The multi-scale joint secondary sub-unit is used to calculate the weighted sum of the optimized fingerprint shallow layer enhancement feature map and the optimized fingerprint structure enhancement feature map to obtain the fingerprint shallow layer-structure significant joint perceptual feature map.

9. The multi-level network identity authentication and access control system according to claim 8, characterized in that: The bidirectional dependency calculation secondary subunit is used to: Calculate the transposed matrix of the fingerprint shallow layer enhancement feature shape reshaping matrix multiplied by the fingerprint structural enhancement feature shape reshaping matrix to obtain a fingerprint shallow layer-structural feature association representation matrix; The fingerprint shallow-structure feature association representation matrix is ​​divided by the scale of the fingerprint structure enhancement feature shape reshaping matrix and then input into the softmax function to obtain the fingerprint shallow-structure feature dependency matrix; Calculate the transposed matrix of the fingerprint shallow layer enhanced feature shape reshaping matrix and multiply it by the fingerprint structure enhanced feature shape reshaping matrix to obtain a fingerprint structure-shallow layer feature association representation matrix; The fingerprint structure-shallow feature association representation matrix is ​​divided by the scale of the fingerprint shallow enhanced feature shape reshaping matrix and then input into the softmax function to obtain the fingerprint structure-shallow feature dependency matrix.

10. The multi-level network identity authentication and access control system according to claim 9, characterized in that: The verification result generating unit comprises: A difference calculation subunit, used to calculate a fingerprint pixel-by-pixel difference matrix between the defogging fingerprint image and the authorized fingerprint image; The fingerprint authentication subunit is used to input the fingerprint pixel-by-pixel difference matrix into the verification result generator based on the classifier to obtain a second-level verification result, wherein the second-level verification result is used to indicate whether the fingerprint authentication is passed.

Images