An identity authentication control method and system based on smart access control and two-ticket linkage

Abstract

The present invention provides an identity authentication control method and system based on the linkage of smart access control and two tickets. The method comprises the following steps: performing image de-cluttering processing on a disassembled image; performing biometric feature extraction on the de-cluttered image, and performing quality screening on biometric features of a human face; performing facial feature extraction on the de-cluttered image; weighted fusion of the screening features with the facial recognition features, and performing identity authentication on the target based on the weighted features to output an identity authentication result; if the identity authentication passes, then the legality of the two-ticket information is verified; if the two-ticket data exists and the two-ticket information is legal, then the smart access control device is controlled to be in an on state; if the two-ticket data does not exist or the identity authentication fails or the two-ticket information is illegal, then the smart access control device is controlled to be in an off state. The present invention can effectively eliminate noise in an image, and can effectively improve the accuracy of facial recognition while reducing the probability of misidentification, thereby improving the security of identity authentication.

Description

Technical Field

[0001] The present invention belongs to the technical field of identity authentication and recognition, and specifically relates to an identity authentication control method and system based on smart access control and two-ticket linkage. Background Art

[0002] Stations usually use a two-ticket system (work ticket and operation ticket) as a means of control to manage on-site operation and maintenance personnel, guide and supervise operation and maintenance personnel to perform related operation and maintenance operations safely and in compliance with regulations. For actual station operation and maintenance operations, operators generally rely on manual inspection of two tickets to enter the station to perform operation and maintenance operations. However, such management is prone to negligence, which in turn affects operational safety. The existing technology also uses facial recognition to identify and authenticate the identity of operators, but in actual situations, due to the influence of external environmental factors or the influence of the operator's clothing, it is easy to cause the actual facial image to be unclear or there is noise in the image, affecting the accuracy of the final identity authentication. Summary of the invention

[0003] In order to solve the above technical problems, the present invention provides an identity authentication control method and system based on smart access control and two-ticket linkage, which is used to solve the technical problems in the prior art.

[0004] On the one hand, the present invention provides the following technical solution, an identity authentication control method based on smart access control and two-ticket linkage, comprising:

[0005] The target video is acquired in real time through the smart access control device, and the target video is disassembled frame by frame to obtain a number of continuous single-frame disassembled images;

[0006] Acquire a face template image, and perform image de-impurity processing on the disassembled image based on the face template image to obtain a de-impurity image;

[0007] Extracting biometric features from the cleaned image to obtain facial biometric features, and performing quality screening on the facial biometric features to obtain screening features;

[0008] Extracting facial features from the cleaned image to obtain facial recognition features;

[0009] The screening feature is weightedly fused with the face recognition feature to obtain a weighted feature, and the target is authenticated based on the weighted feature to output an authentication result;

[0010] If the identity authentication result is that the identity authentication is passed, the two-ticket database of the corresponding time period is indexed based on the identity authentication result and the legitimacy of the two-ticket information is verified according to the location of the smart access control device and the identity authentication result. If there are two-ticket data corresponding to the identity authentication result and the two-ticket information is legal, the smart access control device is controlled to be in the on state. If there are no two-ticket data corresponding to the identity authentication result or the identity authentication result is that the identity authentication fails or the two-ticket information is illegal, the smart access control device is controlled to be in the off state.

[0011] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention first obtains the target video in real time through the smart access control device, disassembles the target video frame by frame to obtain a number of continuous single-frame disassembled images; then obtains the face template image, and performs image de-impurity processing on the disassembled image based on the face template image to obtain a de-impurity image; then performs biometric feature extraction on the de-impurity image to obtain face biometric features, and performs quality screening on the face biometric features to obtain screening features; then performs face feature extraction on the de-impurity image to obtain face recognition features; then performs weighted fusion of the screening features and the face recognition features to obtain weighted features, and performs identity authentication on the target based on the weighted features to output an identity authentication result; finally, if the identity authentication result is that the identity authentication is passed, then the two-ticket database of the corresponding time period is indexed based on the identity authentication result and the location of the smart access control device and the The identity authentication result verifies the legality of the two pieces of information. If there are two pieces of data corresponding to the identity authentication result and the two pieces of information are legal, the smart access control device is controlled to be in an open state. If there are no two pieces of data corresponding to the identity authentication result or the identity authentication result is that the identity authentication fails or the two pieces of information are illegal, the smart access control device is controlled to be in a closed state. The present invention can effectively remove the image due to environmental factors and other noise in the image by removing impurities from the image, and can also remove some debris blocking the face to ensure that the face can be fully and accurately reflected in the image, thereby improving the accuracy of subsequent authentication and recognition. At the same time, the present invention simultaneously extracts facial biometric features and facial recognition features, and weightedly fuses the two features. Through multi-faceted facial feature information, the accuracy of facial recognition can be effectively improved while reducing the probability of misrecognition, thereby improving the security of identity authentication.

[0012] Preferably, the step of performing image de-impurity processing on the disassembled image based on the face template image to obtain the de-impurity image comprises:

[0013] Construct an initial generator and an initial discriminator, input the disassembled image into the initial generator and generate random impurity pixels, and determine the impurity image based on the impurity pixels and the disassembled image :

[0014] ;

[0015] In the formula, Represents the mask corresponding to the random impurity pixel, Representation and Tensors of the same shape, represents element-wise multiplication, Represents a disassembled image;

[0016] The impurity image Paired with the face template image to obtain a first input image pair , for the first input image pair Perform the first de-cluttering process to obtain the first de-cluttered image :

[0017] ;

[0018] In the formula, represents the upsampling operation, represents the downsampling operation, Representation and Paired face template images;

[0019] Generate initial noise , the initial noise is mapped to obtain a mapping feature :

[0020] ;

[0021] In the formula, Represents a nonlinear feature mapping function;

[0022] The mapping feature , the first cleaned image And the impurity image Pairing to obtain the second input image pair , for the second input image pair Perform a second impurity removal process to obtain a second impurity removal image :

[0023] ;

[0024] In the formula, Denotes a decoding operation, Represents contextual attention extraction operation;

[0025] Based on the second impurity removal image Calculating the generator loss and the discriminator loss , and according to the generator loss and the discriminator loss The initial generator and the initial discriminator are updated, and a de-impurified image is output.

[0026] Preferably, the second impurity-removed image Calculating the generator loss and the discriminator loss , and according to the generator loss and the discriminator loss The steps of updating the initial generator and the initial discriminator and outputting the de-impurified image include:

[0027] Based on the second impurity removal image Calculating the generator loss and the discriminator loss :

[0028] ;

[0029] ;

[0030] In the formula, represents the initial generator, represents the image generated by the initial generator, represents the mathematical expectation, , Represent the first hyperparameter and the second hyperparameter respectively, Represents resistance to loss, represents the activation function;

[0031] Based on the generator loss With the discriminator loss Back-propagating the initial generator and the initial discriminator and iteratively updating them alternately to obtain an updated generator and an updated discriminator;

[0032] The update generator and the update discriminator perform a first de-impurity and a second de-impurity on the disassembled image to obtain a de-impurity image.

[0033] Preferably, the step of extracting biometric features from the cleaned image to obtain facial biometric features includes:

[0034] Extract the skin pixel value of the de-impurified image in the RGB space , and determine the standard equation based on the skin pixel value:

[0035] ;

[0036] In the formula, is a standard matrix, a color vector representing skin reflectance, Indicates the skin reflection intensity, represents the static reflection component, Indicates time;

[0037] Solve the standard matrix based on the standard equation , and based on the standard matrix The skin pixel value Perform standardized projection processing to obtain standardized skin pixel values :

[0038] ;

[0039] In the formula, represents the normalized projection axis, represents the skin color direction of the normalized projection, represents the projected color vector, represents the dynamic reflection component, Represents the pulsation intensity in RGB space, Indicates initialization of facial biometric features;

[0040] The standardized skin pixel value Diagonal mapping and projection processing are performed in sequence to obtain projected skin pixel values, and facial biometric features are determined based on the projected skin pixel values.

[0041] Preferably, the standardized skin pixel value The steps of performing diagonal mapping and projection processing in sequence to obtain projected skin pixel values, and determining facial biometric features based on the projected skin pixel values ​​include:

[0042] The standardized skin pixel value Perform diagonal mapping to obtain mapped skin pixel values :

[0043] ;

[0044] In the formula, represents a diagonal mapping matrix;

[0045] The mapped skin pixel value Projection onto orthogonal plane , to get the projected skin pixel value :

[0046] ;

[0047] Through orthogonal planes Project the skin pixel value Orthogonal projection is the first projection signal With the second projection signal , based on the first projection signal With the second projection signal Determine facial biometrics :

[0048] ;

[0049] In the formula, is the standard deviation operator.

[0050] Preferably, the step of performing quality screening on the facial biometric features to obtain screening features comprises:

[0051] Each of the facial biometric features is subjected to window normalization processing to obtain a feature sequence :

[0052] ;

[0053] In the formula, Indicates the first Features, Indicates the first biometric feature of a face Features, represents the mean value of facial biometric features, Indicates the standard deviation of facial biometrics;

[0054] Select a preset template feature sequence , and calculate the Euclidean distance between each feature data in the preset template feature sequence and each feature sequence in the feature sequence, and construct a distance matrix based on the Euclidean distance ;

[0055] From the distance matrix The elements in the first row and first column of the matrix are expanded adjacently until they are expanded to the distance matrix Middle Line The elements of the column and calculate the extension penalty value for each extension :

[0056] ;

[0057] In the formula, Represents the distance matrix The first row and first column of the distance matrix Middle Line The distance between elements of a column, , , Represents the distance matrix Middle Line The elements of the column are extended to Line The elements of the column, Line The elements of the column, Line The expansion penalty value of the elements of the column;

[0058] Determine from the distance matrix The elements in the first row and first column of the distance matrix are expanded to Middle Line All the extended paths of the elements of the column and calculate the matching distance of all the extended paths :

[0059] ;

[0060] In the formula, Indicates the number of expansions. Indicates the first path nodes;

[0061] The facial biometric features whose matching distance is not within the preset range are eliminated, and the facial biometric features whose matching distance is within the preset range are retained to obtain the screening features.

[0062] Preferably, the step of extracting facial features from the cleaned image to obtain facial recognition features comprises:

[0063] Convert the cleaned image into a coefficient matrix in a sparse space With the feature matrix , and calculate the weight matrix based on the coefficient matrix and the feature matrix :

[0064] ;

[0065] In the formula, , , , They represent the weight matrix, the observation matrix of the cleaned image, the coefficient matrix, and the feature matrix. Line The elements of the column, is the regulating factor;

[0066] Based on the weight matrix For the coefficient matrix With the feature matrix Perform iterative update processing to obtain the update coefficient matrix and update the feature matrix :

[0067] ;

[0068] ;

[0069] In the formula, Indicates The updated coefficient matrix after the iteration singular values, Representation Matrix No. singular values, represents the observation matrix of the cleaned image, , , , Respectively represent The updated coefficient matrix after the iteration, The Lagrangian operator after the iteration, The penalty coefficient after the iteration, The updated feature matrix after the iteration, The weight matrix after iterations is Indicates The standard deviation of the singular values ​​of the updated coefficient matrix after iterations, Representation Matrix No. diagonal elements, represents the regularization factor, , Respectively represent sequence After the iteration, the updated feature matrix Line Elements of a column;

[0070] Based on the update coefficient matrix Update the feature matrix with The Lagrangian operator and penalty matrix are iteratively updated to obtain the iterated Lagrangian operator and the penalty coefficient after iteration :

[0071] ;

[0072] ;

[0073] Based on the update coefficient matrix , update the feature matrix , the iterated Lagrangian operator , the penalty matrix after iteration Calculate iterative value :

[0074] ;

[0075] In the formula, represents the two-norm, Indicates The weight matrix after iterations;

[0076] Judgement Iteration value after iterations Is it less than the preset iteration value? Iteration value after iterations is not less than the preset iteration value, the iteration process continues. Iteration value after iterations If it is less than a preset iteration value, the corresponding update coefficient matrix and update feature matrix are output, and the face recognition features are determined based on the output update coefficient matrix and update feature matrix.

[0077] In a second aspect, the present invention provides the following technical solution, an identity authentication control system based on smart access control and two-ticket linkage, the system comprising:

[0078] A disassembly module is used to obtain a target video in real time through a smart access control device, and disassemble the target video frame by frame to obtain a number of continuous single-frame disassembly images;

[0079] A de-impurity module, used for acquiring a face template image, and performing image de-impurity processing on the disassembled image based on the face template image to obtain a de-impurity image;

[0080] A first extraction module is used to extract biometric features from the cleaned image to obtain facial biometric features, and perform quality screening on the facial biometric features to obtain screening features;

[0081] A second extraction module is used to extract facial features from the cleaned image to obtain facial recognition features;

[0082] An authentication module, used for weightedly fusing the screening feature with the face recognition feature to obtain a weighted feature, performing identity authentication on the target based on the weighted feature, and outputting an identity authentication result;

[0083] The access control module is used to, if the identity authentication result is that the identity authentication is passed, index the two-ticket database of the corresponding time period based on the identity authentication result and verify the legitimacy of the two-ticket information according to the location of the smart access control device and the identity authentication result; if there are two-ticket data corresponding to the identity authentication result and the two-ticket information is legal, the smart access control device is controlled to be in an open state; if there are no two-ticket data corresponding to the identity authentication result or the identity authentication result is that the identity authentication is failed or the two-ticket information is illegal, the smart access control device is controlled to be in a closed state.

[0084] In the third aspect, the present invention provides the following technical solution: a computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor; when the processor executes the computer program, the identity authentication control method based on smart access control and two-ticket linkage as described above is implemented.

[0085] In a fourth aspect, the present invention provides the following technical solution: a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the identity authentication control method based on the linkage between smart access control and two tickets as described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0087] Figure 1 A flow chart of an identity authentication control method based on smart access control and two-ticket linkage provided in Embodiment 1 of the present invention;

[0088] Figure 2 A structural block diagram of an identity authentication control system based on smart access control and two-ticket linkage provided in the second embodiment of the present invention;

[0089] Figure 3 A schematic diagram of the hardware structure of a computer provided in another embodiment of the present invention.

[0090] The embodiments of the present invention will be further described below with reference to the accompanying drawings. DETAILED DESCRIPTION

[0091] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the embodiments of the present invention, and should not be construed as limiting the present invention.

[0092] In the description of the embodiments of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0093] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0094] In the embodiments of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the embodiments of the present invention can be understood according to specific circumstances.

[0095] Embodiment 1

[0096] In the first embodiment of the present invention, Figure 1 As shown, an identity authentication control method based on smart access control and two-ticket linkage includes:

[0097] S1. Obtain a target video in real time through a smart access control device, and disassemble the target video frame by frame to obtain several consecutive single-frame disassembled images;

[0098] Specifically, the camera device attached to the smart access control device can capture the target image within a certain range in front of the smart access control device, and split the captured video frame by frame according to the time series to obtain several connected disassembled images. In actual situations, the camera device can also be controlled to capture the target image according to a preset time period and directly output the target image.

[0099] S2, obtaining a face template image, and performing image de-impurity processing on the disassembled image based on the face template image to obtain a de-impurity image;

[0100] Specifically, in actual situations, affected by the shooting equipment, environmental factors and the target's clothing, such as safety helmets, safety helmets are usually required to be worn when performing maintenance or other operations at the site. However, since different workers wear safety helmets in different ways, the face in the final captured image may be obscured by the safety helmet, which may cause the actual disassembly image to contain more noise, or even partially obstructed and blurred face. Therefore, in this step, the features of the face template image are learned and the disassembly image is dedusted according to the features of the face template image to obtain a clear dedusted image.

[0101] Wherein, the step S2 comprises:

[0102] S21, constructing an initial generator and an initial discriminator, inputting the disassembled image into the initial generator and generating random impurity pixels, and determining an impurity image based on the impurity pixels and the disassembled image :

[0103] ;

[0104] In the formula, Represents the mask corresponding to the random impurity pixel, Representation and Tensors of the same shape, represents element-wise multiplication, Represents a disassembled image;

[0105] Specifically, the random impurity pixel point may be a single-channel mask, the pixel value of the blocked portion is 1, and the pixel value of the unblocked portion is 0.

[0106] S22, the impurity image Paired with the face template image to obtain a first input image pair , for the first input image pair Perform the first de-cluttering process to obtain the first de-cluttered image :

[0107] ;

[0108] In the formula, represents the upsampling operation, represents the downsampling operation, Representation and Paired face template images;

[0109] Specifically, through upsampling processing, the intermediate latent codes and feature pixel information in the face template image can be learned, the upsampling processing can decode high-level features, and the image is restored for the first time according to the learned intermediate latent codes and feature pixel information, that is, the first de-noising processing. After the first de-noising, most of the noise can be removed, but after the first de-noising, the pixels of the occluded part are still not clear enough, so a second de-noising processing is needed to improve the clarity of the image.

[0110] S23, generate initial noise , the initial noise is mapped to obtain a mapping feature :

[0111] ;

[0112] In the formula, Represents a nonlinear feature mapping function.

[0113] S24, mapping the features , the first cleaned image And the impurity image Pairing to obtain the second input image pair , for the second input image pair Perform a second impurity removal process to obtain a second impurity removal image :

[0114] ;

[0115] In the formula, Denotes a decoding operation, Represents contextual attention extraction operation;

[0116] Specifically, the initial noise after mapping can learn the image style features of the face template image and the first de-noising image, and through contextual attention extraction processing, it can simulate the randomness in the middle of the occluded part, and at the same time, it can also adjust the deep feature representation suitable for the first de-noising stage.

[0117] S25, based on the second impurity-removed image Calculating the generator loss and the discriminator loss , and according to the generator loss and the discriminator loss Updating the initial generator and the initial discriminator, and outputting a cleaned image;

[0118] Wherein, the step S25 comprises:

[0119] S251: Based on the second impurity-removed image Calculating the generator loss and the discriminator loss :

[0120] ;

[0121] ;

[0122] In the formula, represents the initial generator, represents the image generated by the initial generator, represents the mathematical expectation, , Represent the first hyperparameter and the second hyperparameter respectively, Represents resistance to loss, Represents the activation function.

[0123] S252, based on the generator loss With the discriminator loss The initial generator and the initial discriminator are back-propagated and iteratively updated alternately to obtain an updated generator and an updated discriminator.

[0124] S253, performing a first impurity removal and a second impurity removal on the disassembled image by the update generator and the update discriminator to obtain an impurity-removed image;

[0125] Specifically, the image generated by the updated generator after training and updating can simulate the real image, and after being judged by the discriminator, it can reach the point where it is indistinguishable from the real image. Therefore, by inputting the face template image and the disassembled image into the updated generator and the updated discriminator, the final clear and noise-free cleaned image can be output.

[0126] S3, extracting biometric features from the cleaned image to obtain facial biometric features, and performing quality screening on the facial biometric features to obtain screening features;

[0127] Wherein, the step S3 comprises:

[0128] S311, extracting the skin pixel value of the de-impurified image in the RGB space , and determine the standard equation based on the skin pixel value:

[0129] ;

[0130] In the formula, is a standard matrix, a color vector representing skin reflectance, Indicates the skin reflection intensity, represents the static reflection component, Indicates time;

[0131] Specifically, a face image not only contains the characteristic information of the face, but also includes the photoelectric oscillation information of the target skin color after being exposed to light, which can also be used for specific identity authentication of the person.

[0132] S312, solving the standard matrix based on the standard equation , and based on the standard matrix The skin pixel value Perform standardized projection processing to obtain standardized skin pixel values :

[0133] ;

[0134] In the formula, represents the normalized projection axis, represents the skin color direction of the normalized projection, represents the projected color vector, represents the dynamic reflection component, Represents the pulsation intensity in RGB space, Indicates initialization of facial biometric features;

[0135] Specifically, after standardized projection, the light intensity on the projection axis changes with time, which improves the flexibility of feature acquisition and reduces the sensitivity to prior information.

[0136] S313: normalizing the skin pixel value Performing diagonal mapping and projection processing in sequence to obtain projected skin pixel values, and determining facial biometric features based on the projected skin pixel values;

[0137] Wherein, the step S313 includes:

[0138] S3131, the standardized skin pixel value Perform diagonal mapping to obtain mapped skin pixel values :

[0139] ;

[0140] In the formula, Represents a diagonal mapping matrix.

[0141] Specifically, by performing diagonal mapping processing, the image can be automatically white balanced, thereby reducing the impact of the environment on the extraction of facial biometric features.

[0142] S3132, mapping the skin pixel value Projection onto orthogonal plane , to get the projected skin pixel value :

[0143] ;

[0144] Specifically, after diagonal mapping, the reflection component of the mirror It should be projected in the direction of white light, so its specific value can be ignored. Therefore, in the orthogonal projection process, there is no need to consider the reflection component of the mirror

[0145] S3133, through the orthogonal plane Project the skin pixel value Orthogonal projection is the first projection signal With the second projection signal , based on the first projection signal With the second projection signal Determine facial biometrics :

[0146] ;

[0147] In the formula, is the standard deviation operator;

[0148] Specifically, after projection onto the orthogonal plane, and since the orthogonal plane is a projection distance of 2×3 composed of a projection axis and is defined as a plane in the RGB space, and since the change in light motion occurs in the same phase, after orthogonal projection, a first projection signal with positive and negative phases is formed With the second projection signal , through the first projection signal With the second projection signal Eliminates the impact of light movement and improves the balance of light intensity, thereby determining accurate facial biometric features. .

[0149] Wherein, the step S3 further includes:

[0150] S321, performing window normalization processing on each of the facial biometric features to obtain a feature sequence :

[0151] ;

[0152] In the formula, Indicates the first Features, Indicates the first biometric feature of a face Features, represents the mean value of facial biometric features, Indicates the standard deviation of facial biometrics;

[0153] Specifically, after obtaining the facial biometric features, due to the influence of human body movement and lighting, some interfered feature signals may exist in the finally extracted facial biometric features. Therefore, in order to improve the accuracy of face recognition, they need to be eliminated.

[0154] S322, select a preset template feature sequence , and calculate the Euclidean distance between each feature data in the preset template feature sequence and each feature sequence in the feature sequence, and construct a distance matrix based on the Euclidean distance .

[0155] S323, from the distance matrix The elements in the first row and first column of the matrix are expanded adjacently until they are expanded to the distance matrix Middle Line The elements of the column and calculate the extension penalty value for each extension :

[0156] ;

[0157] In the formula, Represents the distance matrix The first row and first column of the distance matrix Middle Line The distance between elements of a column, , , Represents the distance matrix Middle Line The elements of the column are extended to Line The elements of the column, Line The elements of the column, Line The expansion penalty value for the elements of the column.

[0158] S324, determine from the distance matrix The elements in the first row and first column of the distance matrix are expanded to Middle Line All the extended paths of the elements of the column and calculate the matching distance of all the extended paths :

[0159] ;

[0160] In the formula, Indicates the number of expansions. Indicates the first Path nodes.

[0161] S325, eliminating facial biometric features whose matching distances are not within a preset range, and retaining facial biometric features whose matching distances are within the preset range, to obtain screening features;

[0162] Specifically, there are several expansion paths. For the required features, during the expansion process, the matching distance should be as small as possible, and the distance between the features and the preset template feature sequence should also be as small as possible. Therefore, facial biometric features within a preset range are selected for retention. In actual situations, the matching distance corresponding to each facial biometric feature is sorted in ascending order. In this embodiment, only the top 15% of the corresponding facial biometric features are selected as the final screening feature output.

[0163] S4, extracting facial features from the cleaned image to obtain facial recognition features;

[0164] Specifically, step S4 includes:

[0165] S41, converting the cleaned image into a coefficient matrix in a sparse space With the feature matrix , and calculate the weight matrix based on the coefficient matrix and the feature matrix :

[0166] ;

[0167] In the formula, , , , They represent the weight matrix, the observation matrix of the cleaned image, the coefficient matrix, and the feature matrix. Line The elements of the column, is the adjustment factor.

[0168] S42, based on the weight matrix For the coefficient matrix With the feature matrix Perform iterative update processing to obtain the update coefficient matrix and update the feature matrix :

[0169] ;

[0170] ;

[0171] In the formula, Indicates The updated coefficient matrix after the iteration singular values, Representation Matrix No. singular values, represents the observation matrix of the cleaned image, , , , Respectively represent The updated coefficient matrix after the iteration, The Lagrangian operator after the iteration, The penalty coefficient after the iteration, The updated feature matrix after the iteration, The weight matrix after iterations is Indicates The standard deviation of the singular values ​​of the updated coefficient matrix after iterations, Representation Matrix No. diagonal elements, represents the regularization factor, , Respectively represent sequence After the iteration, the updated feature matrix Line Elements of a column;

[0172] Specifically, in the actual iteration process, the weight matrix It will also change with the changes of the new coefficient matrix and the updated feature matrix. In each round of iteration, it is iteratively updated according to the formula in step S41.

[0173] S43, based on the update coefficient matrix Update the feature matrix with The Lagrangian operator and penalty matrix are iteratively updated to obtain the iterated Lagrangian operator and the penalty coefficient after iteration :

[0174] ;

[0175] .

[0176] S44, based on the update coefficient matrix , update the feature matrix , the iterated Lagrangian operator , the penalty matrix after iteration Calculate iterative value :

[0177] ;

[0178] In the formula, represents the two-norm, Indicates The weight matrix after iterations;

[0179] S45. Judgment No. Iteration value after iterations Is it less than the preset iteration value? Iteration value after iterations is not less than the preset iteration value, the iteration process continues. Iteration value after iterations If the value is less than the preset iteration value, the corresponding update coefficient matrix and update feature matrix are output, and the face recognition features are determined based on the output update coefficient matrix and update feature matrix;

[0180] Specifically, by iteratively updating the update coefficient matrix and the update feature matrix, occlusion, noise and light pollution in the image can be effectively removed, further improving the accuracy of face recognition.

[0181] S5, weighting and fusing the screening feature with the face recognition feature to obtain a weighted feature, and authenticating the target based on the weighted feature to output an authentication result;

[0182] Specifically, in this step, the weighted features ,in, , Represent the first weight and the second weight respectively, , represents the screening feature, Represents face recognition features. In actual situations, the values ​​of the first weight and the second weight can be based on the actual shooting situation. If the face recognition feature information can be fully extracted, the second weight is greater than the first weight. If the screening feature can be fully extracted, the second weight is less than the first weight. Specifically, after obtaining the weighted features, a search is performed in the locally stored feature database to calculate the similarity between the weighted features and the feature data in the local feature database. If the similarity is greater than a preset value, the identity authentication result is that the identity authentication is passed. If the similarity is not greater than the preset value, the identity authentication structure is that the identity authentication is failed.

[0183] S6. If the identity authentication result is that the identity authentication is passed, then based on the identity authentication result, the two-ticket database of the corresponding time period is indexed and the legitimacy of the two-ticket information is verified according to the location of the smart access control device and the identity authentication result. If there is two-ticket data corresponding to the identity authentication result and the two-ticket information is legal, then the smart access control device is controlled to be in an open state; if there is no two-ticket data corresponding to the identity authentication result or the identity authentication result is that the identity authentication is failed or the two-ticket information is illegal, then the smart access control device is controlled to be in a closed state;

[0184] Specifically, if there are two tickets of data corresponding to the identity authentication result and the identity authentication is passed and the location information of the two tickets can match the location information of the smart access control device and the two tickets are legal, the smart access control device can be controlled to be in an open state and the target is given access permission. If there are no two tickets of data corresponding to the identity authentication result or the identity authentication is not passed or the location information of the two tickets cannot match the location information of the smart access control device or the two tickets are illegal, as long as any of the above conditions is not met, the smart access control device is controlled to be in a closed state and the target is not given access permission. If the target performs the authentication process for a long time and multiple times and the authentication fails, the smart access control device will alarm and upload the alarm information to the cloud;

[0185] Among them, the legality of the two tickets is determined by the location of the smart access control and the corresponding identity authentication results. After the two tickets are issued, the two ticket information contains the location information of the smart access control and the corresponding identity information. If when the staff passes through a smart access control device, if the two ticket information does not contain the location information of the smart access control device currently required to pass through, it means that the two tickets are illegal, and if the identity authentication results of the staff are different, the two ticket information is also illegal, and vice versa.

[0186] The identity authentication control method based on the linkage of smart access control and two tickets provided in the first embodiment of the present invention first obtains the target video in real time through the smart access control device, disassembles the target video frame by frame to obtain a number of continuous single-frame disassembled images; then obtains the face template image, and performs image de-impurity processing on the disassembled image based on the face template image to obtain a de-impurity image; then performs biometric feature extraction on the de-impurity image to obtain face biometric features, and performs quality screening on the face biometric features to obtain screening features; then performs face feature extraction on the de-impurity image to obtain face recognition features; then performs weighted fusion of the screening features and the face recognition features to obtain weighted features, and performs identity authentication on the target based on the weighted features to output the identity authentication result; finally, if the identity authentication result is that the identity authentication is passed, the two-ticket database of the corresponding time period is indexed based on the identity authentication result and the result is obtained according to the smart access control device. The position of the image and the identity authentication result verify the legitimacy of the two pieces of information. If there are two pieces of data corresponding to the identity authentication result and the two pieces of information are legal, the smart access control device is controlled to be in an open state. If there are no two pieces of data corresponding to the identity authentication result or the identity authentication result is that the identity authentication fails or the two pieces of information are illegal, the smart access control device is controlled to be in a closed state. The present invention can effectively remove the image due to environmental factors and other noise in the image by removing impurities from the image, and can also remove some debris blocking the face to ensure that the face can be completely and accurately reflected in the image, thereby improving the accuracy of subsequent authentication and recognition. At the same time, the present invention simultaneously extracts facial biometric features and facial recognition features, and weightedly fuses the two features. Through multi-faceted facial feature information, the accuracy of facial recognition can be effectively improved while reducing the probability of misrecognition, thereby improving the security of identity authentication.

[0187] Embodiment 2

[0188] like Figure 2 As shown, in the second embodiment of the present invention, an identity authentication control system based on smart access control and two-ticket linkage is provided, and the system includes:

[0189] The disassembly module 1 is used to obtain the target video in real time through the smart access control device, and disassemble the target video frame by frame to obtain a number of continuous single-frame disassembly images;

[0190] The impurity removal module 2 is used to obtain a face template image, and perform image impurity removal processing on the disassembled image based on the face template image to obtain an impurity-removed image;

[0191] The first extraction module 3 is used to extract biometric features from the cleaned image to obtain facial biometric features, and perform quality screening on the facial biometric features to obtain screening features;

[0192] The second extraction module 4 is used to extract facial features from the cleaned image to obtain facial recognition features;

[0193] An authentication module 5 is used to perform weighted fusion of the screening feature and the face recognition feature to obtain a weighted feature, perform identity authentication on the target based on the weighted feature, and output an identity authentication result;

[0194] The access control module 6 is used for, if the identity authentication result is that the identity authentication is passed, indexing in the two-ticket database of the corresponding time period based on the identity authentication result and verifying the legitimacy of the two-ticket information according to the location of the smart access control device and the identity authentication result; if there is two-ticket data corresponding to the identity authentication result and the two-ticket information is legal, then the smart access control device is controlled to be in an open state; if there is no two-ticket data corresponding to the identity authentication result or the identity authentication result is that the identity authentication is not passed or the two-ticket information is illegal, then the smart access control device is controlled to be in a closed state;

[0195] The impurity removal module 2 comprises:

[0196] A construction submodule is used to construct an initial generator and an initial discriminator, input the disassembled image into the initial generator and generate random impurity pixels, and determine the impurity image based on the impurity pixels and the disassembled image. :

[0197] ;

[0198] In the formula, Represents the mask corresponding to the random impurity pixel, Representation and Tensors of the same shape, represents element-wise multiplication, Represents a disassembled image;

[0199] A pairing submodule is used to pair the impurity image Paired with the face template image to obtain a first input image pair , for the first input image pair Perform the first de-cluttering process to obtain the first de-cluttered image :

[0200] ;

[0201] In the formula, represents the upsampling operation, represents the downsampling operation, Representation and Paired face template images;

[0202] Generate submodule for generating initial noise , the initial noise is mapped to obtain a mapping feature :

[0203] ;

[0204] In the formula, Represents a nonlinear feature mapping function;

[0205] The impurity removal module is used to map the features , the first cleaned image And the impurity image Pairing to obtain the second input image pair , for the second input image pair Perform a second impurity removal process to obtain a second impurity removal image :

[0206] ;

[0207] In the formula, Denotes a decoding operation, Represents contextual attention extraction operation;

[0208] A loss submodule for removing impurities from the image based on the second Calculating the generator loss and the discriminator loss , and according to the generator loss and the discriminator loss The initial generator and the initial discriminator are updated, and a de-impurified image is output.

[0209] The loss submodule includes:

[0210] A loss calculation unit is used to calculate the loss of the image based on the second impurity removal image. Calculating the generator loss and the discriminator loss :

[0211] ;

[0212] ;

[0213] In the formula, represents the initial generator, represents the image generated by the initial generator, represents the mathematical expectation, , Represent the first hyperparameter and the second hyperparameter respectively, Represents resistance to loss, represents the activation function;

[0214] Propagation unit for the generator loss With the discriminator loss Back-propagating the initial generator and the initial discriminator and iteratively updating them alternately to obtain an updated generator and an updated discriminator;

[0215] The impurity removal unit is used to perform a first impurity removal and a second impurity removal on the disassembled image through the update generator and the update discriminator to obtain an impurity-removed image.

[0216] The first extraction module comprises:

[0217] The equation determines the submodule, which is used to extract the skin pixel value of the de-impurified image in the RGB space. , and determine the standard equation based on the skin pixel value:

[0218] ;

[0219] In the formula, is a standard matrix, a color vector representing skin reflectance, Indicates the skin reflection intensity, represents the static reflection component, Indicates time;

[0220] A solver submodule for solving the standard matrix based on the standard equation , and based on the standard matrix The skin pixel value Perform standardized projection processing to obtain standardized skin pixel values :

[0221] ;

[0222] In the formula, represents the normalized projection axis, represents the skin color direction of the normalized projection, represents the projected color vector, represents the dynamic reflection component, Represents the pulsation intensity in RGB space, Indicates initialization of facial biometric features;

[0223] A mapping projection submodule is used to normalize the skin pixel value Diagonal mapping and projection processing are performed in sequence to obtain projected skin pixel values, and facial biometric features are determined based on the projected skin pixel values.

[0224] The mapping projection submodule includes:

[0225] A mapping unit for converting the standardized skin pixel value Perform diagonal mapping to obtain mapped skin pixel values :

[0226] ;

[0227] In the formula, represents a diagonal mapping matrix;

[0228] A projection unit is used to project the skin pixel value Projection onto orthogonal plane , to get the projected skin pixel value :

[0229] ;

[0230] Feature output unit for passing through orthogonal planes Project the skin pixel value Orthogonal projection is the first projection signal With the second projection signal , based on the first projection signal With the second projection signal Determine facial biometrics :

[0231] ;

[0232] In the formula, is the standard deviation operator.

[0233] The first extraction module 3 also includes:

[0234] The normalization submodule is used to perform window normalization processing on each of the facial biometric features to obtain a feature sequence :

[0235] ;

[0236] In the formula, Indicates the first Features, Indicates the first biometric feature of a face Features, represents the mean value of facial biometric features, Indicates the standard deviation of facial biometrics;

[0237] Distance calculation submodule, used to select preset template feature sequence , and calculate the Euclidean distance between each feature data in the preset template feature sequence and each feature sequence in the feature sequence, and construct a distance matrix based on the Euclidean distance ;

[0238] Extension submodule for extracting the distance matrix The elements in the first row and first column of the matrix are expanded adjacently until they are expanded to the distance matrix Middle Line The elements of the column and calculate the extension penalty value for each extension :

[0239] ;

[0240] In the formula, Represents the distance matrix The first row and first column of the distance matrix Middle Line The distance between elements of a column, , , Represents the distance matrix Middle Line The elements of the column are extended to Line The elements of the column, Line The elements of the column, Line The expansion penalty value of the elements of the column;

[0241] The path submodule is used to determine the distance matrix The elements in the first row and first column of the distance matrix are expanded to Middle Line All the extended paths of the elements of the column and calculate the matching distance of all the extended paths :

[0242] ;

[0243] In the formula, Indicates the number of expansions. Indicates the first path nodes;

[0244] The elimination submodule is used to eliminate the facial biometric features whose matching distance is not within the preset range, and retain the facial biometric features whose matching distance is within the preset range to obtain the screening features.

[0245] The second extraction module 4 comprises:

[0246] A matrix submodule, used to convert the cleaned image into a coefficient matrix in a sparse space With the feature matrix , and calculate the weight matrix based on the coefficient matrix and the feature matrix :

[0247] ;

[0248] In the formula, , , , They represent the weight matrix, the observation matrix of the cleaned image, the coefficient matrix, and the feature matrix. Line The elements of the column, is the regulating factor;

[0249] The first iterative submodule is used to For the coefficient matrix With the feature matrix Perform iterative update processing to obtain the update coefficient matrix and update the feature matrix :

[0250] ;

[0251] ;

[0252] In the formula, Indicates The updated coefficient matrix after the iteration singular values, Representation Matrix No. singular values, represents the observation matrix of the cleaned image, , , , Respectively represent The updated coefficient matrix after the iteration, The Lagrangian operator after the iteration, The penalty coefficient after the iteration, The updated feature matrix after the iteration, The weight matrix after iterations is Indicates The standard deviation of the singular values ​​of the updated coefficient matrix after iterations, Representation Matrix No. diagonal elements, represents the regularization factor, , Respectively represent sequence After the iteration, the updated feature matrix Line Elements of a column;

[0253] The second iterative submodule is used to update the coefficient matrix based on the Update the feature matrix with The Lagrangian operator and penalty matrix are iteratively updated to obtain the iterated Lagrangian operator and the penalty coefficient after iteration :

[0254] ;

[0255] ;

[0256] Iteration value submodule, used to update the coefficient matrix based on , update the feature matrix , the iterated Lagrangian operator , the penalty matrix after iteration Calculate iterative value :

[0257] ;

[0258] In the formula, represents the two-norm, Indicates The weight matrix after iterations;

[0259] The judgment submodule is used to judge the Iteration value after iterations Is it less than the preset iteration value? Iteration value after iterations is not less than the preset iteration value, the iteration process continues. Iteration value after iterations If it is less than a preset iteration value, the corresponding update coefficient matrix and update feature matrix are output, and the face recognition features are determined based on the output update coefficient matrix and update feature matrix.

[0260] In some other embodiments of the present invention, the embodiments of the present invention provide the following technical solutions: a computer, comprising a memory 102, a processor 101, and a computer program stored in the memory 102 and executable on the processor 101; when the processor 101 executes the computer program, the identity authentication control method based on the linkage between smart access control and two tickets as described above is implemented.

[0261] Specifically, the processor 101 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiment of the present invention.

[0262] Among them, the memory 102 may include a large-capacity memory for data or instructions. By way of example and not limitation, the memory 102 may include a hard disk drive (HDD), a floppy disk drive, a solid state drive (SSD), a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 102 may include a removable or non-removable (or fixed) medium. Where appropriate, the memory 102 may be inside or outside the data processing device. In a specific embodiment, the memory 102 is a non-volatile memory. In a specific embodiment, the memory 102 includes a read-only memory (ROM) and a random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically alterable ROM (EAROM) or a flash memory (FLASH), or a combination of two or more of these. Under appropriate circumstances, the RAM can be a static random access memory (SRAM) or a dynamic random access memory (DRAM), wherein the DRAM can be a fast page mode dynamic random access memory (FPMDRAM), an extended data output dynamic random access memory (EDODRAM), a synchronous dynamic random access memory (SDRAM), etc.

[0263] The memory 102 may be used to store or cache various data files that need to be processed and / or used for communication, as well as possible computer program instructions executed by the processor 101 .

[0264] The processor 101 implements the above-mentioned identity authentication control method based on the linkage between smart access control and two tickets by reading and executing the computer program instructions stored in the memory 102.

[0265] In some embodiments, the computer may further include a communication interface 103 and a bus 100. Figure 3 As shown, the processor 101, the memory 102, and the communication interface 103 are connected via a bus 100 and communicate with each other.

[0266] The communication interface 103 is used to implement communication between the modules, devices, units and / or equipment in the embodiment of the present invention. The communication interface 103 can also implement data communication with other components such as: external devices, image / data acquisition equipment, databases, external storage, and image / data processing workstations.

[0267] The bus 100 includes hardware, software or both, and couples the components of the computer device to each other. The bus 100 includes but is not limited to at least one of the following: a data bus, an address bus, a control bus, an expansion bus, and a local bus. By way of example and not limitation, bus 100 may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses or a combination of two or more of these. Where appropriate, bus 100 may include one or more buses. Although embodiments of the present invention describe and illustrate a particular bus, the present invention contemplates any suitable bus or interconnect.

[0268] The computer can execute the identity authentication control method based on the linkage of smart access control and two tickets of the present invention based on the acquisition of the identity authentication control system based on the linkage of smart access control and two tickets, thereby realizing the identity authentication control based on the linkage of smart access control and two tickets.

[0269] In some further embodiments of the present invention, in combination with the above-mentioned identity authentication control method based on smart access control and two-ticket linkage, the embodiments of the present invention provide the following technical solutions, a storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned identity authentication control method based on smart access control and two-ticket linkage is implemented.

[0270] Those skilled in the art will appreciate that the logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, "computer-readable medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0271] More specific examples of readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0272] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0273] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0274] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. An identity authentication control method based on smart access control and two-ticket linkage, characterized in that: include: The target video is acquired in real time through the smart access control device, and the target video is disassembled frame by frame to obtain a number of continuous single-frame disassembled images; Acquire a face template image, and perform image de-impurity processing on the disassembled image based on the face template image to obtain a de-impurity image; Extracting biometric features from the cleaned image to obtain facial biometric features, and performing quality screening on the facial biometric features to obtain screening features; Extracting facial features from the cleaned image to obtain facial recognition features; The screening feature is weightedly fused with the face recognition feature to obtain a weighted feature, and the target is authenticated based on the weighted feature to output an authentication result; If the identity authentication result is that the identity authentication is passed, then the two-ticket database of the corresponding time period is indexed based on the identity authentication result and the legitimacy of the two-ticket information is verified according to the location of the smart access control device and the identity authentication result. If there are two-ticket data corresponding to the identity authentication result and the two-ticket information is legal, then the smart access control device is controlled to be in an open state; if there are no two-ticket data corresponding to the identity authentication result or the identity authentication result is that the identity authentication is failed or the two-ticket information is illegal, then the smart access control device is controlled to be in a closed state; The step of performing image de-impurity processing on the disassembled image based on the face template image to obtain a de-impurity image comprises: Construct an initial generator and an initial discriminator, input the disassembled image into the initial generator and generate random impurity pixels, and determine the impurity image based on the impurity pixels and the disassembled image : ； In the formula, Represents the mask corresponding to the random impurity pixel, Representation and Tensors of the same shape, represents element-wise multiplication, Represents a disassembled image; The impurity image Paired with the face template image to obtain a first input image pair , for the first input image pair Perform the first de-cluttering process to obtain the first de-cluttered image : ； In the formula, represents the upsampling operation, represents the downsampling operation, Representation and Paired face template images; Generate initial noise , the initial noise is mapped to obtain a mapping feature : ； In the formula, Represents a nonlinear feature mapping function; The mapping feature , the first cleaned image And the impurity image Pairing is performed to obtain a second input image pair , for the second input image pair Perform a second impurity removal process to obtain a second impurity removal image : ； In the formula, Denotes a decoding operation, Represents contextual attention extraction operation; Based on the second impurity-removed image Calculating the generator loss and the discriminator loss , and according to the generator loss and the discriminator loss Updating the initial generator and the initial discriminator, and outputting a cleaned image; based on the second impurity-removed image Calculating the generator loss and the discriminator loss , and according to the generator loss and the discriminator loss The steps of updating the initial generator and the initial discriminator and outputting the de-impurified image include: Based on the second impurity-removed image Calculating the generator loss and the discriminator loss : ； ； In the formula, represents the initial generator, represents the image generated by the initial generator, represents the mathematical expectation, , Represent the first hyperparameter and the second hyperparameter respectively, Represents resistance to loss, represents the activation function; Based on the generator loss With the discriminator loss Back-propagating the initial generator and the initial discriminator and iteratively updating them alternately to obtain an updated generator and an updated discriminator; The update generator and the update discriminator perform a first de-impurity and a second de-impurity on the disassembled image to obtain a de-impurity image.

2. According to claim 1, the identity authentication control method based on smart access control and two-ticket linkage is characterized in that: The step of extracting biometric features from the cleaned image to obtain facial biometric features comprises: Extract the skin pixel value of the de-impurified image in the RGB space , and determine the standard equation based on the skin pixel value: ； In the formula, is a standard matrix, a color vector representing skin reflectance, Indicates the skin reflection intensity, represents the static reflection component, Indicates time; Solve the standard matrix based on the standard equation , and based on the standard matrix The skin pixel value Perform standardized projection processing to obtain standardized skin pixel values : ； In the formula, represents the normalized projection axis, represents the skin color direction of the normalized projection, represents the projected color vector, represents the dynamic reflection component, Represents the pulsation intensity in RGB space, Indicates initialization of facial biometric features; The standardized skin pixel value Diagonal mapping and projection processing are performed in sequence to obtain projected skin pixel values, and facial biometric features are determined based on the projected skin pixel values.

3. The identity authentication control method based on smart access control and two-ticket linkage according to claim 2 is characterized in that: The standardized skin pixel value The steps of performing diagonal mapping and projection processing in sequence to obtain projected skin pixel values, and determining facial biometric features based on the projected skin pixel values ​​include: The standardized skin pixel value Perform diagonal mapping to obtain mapped skin pixel values : ； In the formula, represents a diagonal mapping matrix; The mapped skin pixel value Projection onto orthogonal plane , to get the projected skin pixel value : ； Through orthogonal planes Project the skin pixel value Orthogonal projection is the first projection signal With the second projection signal , based on the first projection signal With the second projection signal Determine facial biometrics : ； In the formula, is the standard deviation operator.

4. The identity authentication control method based on smart access control and two-ticket linkage according to claim 3 is characterized in that: The step of performing quality screening on the facial biometric features to obtain screening features comprises: Each of the facial biometric features is subjected to window normalization processing to obtain a feature sequence : ； In the formula, Indicates the first Features, Indicates the first biometric feature of a face Features, represents the mean value of facial biometric features, Indicates the standard deviation of facial biometrics; Select a preset template feature sequence , and calculate the Euclidean distance between each feature data in the preset template feature sequence and each feature sequence in the feature sequence, and construct a distance matrix based on the Euclidean distance ; From the distance matrix The elements in the first row and first column of the matrix are expanded adjacently until they are expanded to the distance matrix Middle Line The elements of the column and calculate the extension penalty value for each extension : ； In the formula, Represents the distance matrix The first row and first column of the element to the distance matrix Middle Line The distance between elements of a column, , , Represents the distance matrix Middle Line The elements of the column are extended to Line The elements of the column, Line The elements of the column, Line The expansion penalty value of the elements of the column; Determine from the distance matrix The elements in the first row and first column of the distance matrix are expanded to Middle Line All the extended paths of the elements of the column and calculate the matching distance of all the extended paths : ； In the formula, Indicates the number of expansions. Indicates the first path nodes; The facial biometric features whose matching distance is not within the preset range are eliminated, and the facial biometric features whose matching distance is within the preset range are retained to obtain the screening features.

5. According to claim 1, the identity authentication control method based on smart access control and two-ticket linkage is characterized in that: The step of extracting facial features from the cleaned image to obtain facial recognition features comprises: Convert the cleaned image into a coefficient matrix in a sparse space With the feature matrix , and calculate the weight matrix based on the coefficient matrix and the feature matrix : ； In the formula, , , , They represent the weight matrix, the observation matrix of the cleaned image, the coefficient matrix, and the feature matrix. Line The elements of the column, is the regulating factor; Based on the weight matrix For the coefficient matrix With the feature matrix Perform iterative update processing to obtain the update coefficient matrix and update the feature matrix : ； ； In the formula, Indicates The updated coefficient matrix after the iteration singular values, Representation Matrix No. singular values, represents the observation matrix of the cleaned image, , , , Respectively represent The updated coefficient matrix after the iteration, The Lagrangian operator after the iteration, The penalty coefficient after the iteration, The updated feature matrix after the iteration, The weight matrix after iterations is Indicates The standard deviation of the singular values ​​of the updated coefficient matrix after iterations, Representation Matrix No. diagonal elements, represents the regularization factor, , Respectively represent sequence After the iteration, the updated feature matrix Line Elements of a column; Based on the update coefficient matrix Update the feature matrix with The Lagrangian operator and penalty matrix are iteratively updated to obtain the iterated Lagrangian operator and the penalty coefficient after iteration : ； ； Based on the update coefficient matrix , update the feature matrix , the iterated Lagrangian operator , the penalty matrix after iteration Calculate iterative value : ； In the formula, represents the two-norm, Indicates The weight matrix after iterations; Judgement Iteration value after iterations Is it less than the preset iteration value? Iteration value after iterations is not less than the preset iteration value, the iteration process continues. Iteration value after iterations If it is less than a preset iteration value, the corresponding update coefficient matrix and update feature matrix are output, and the face recognition features are determined based on the output update coefficient matrix and update feature matrix.

6. An identity authentication control system based on smart access control and two-ticket linkage, the system adopts the identity authentication control method based on smart access control and two-ticket linkage as claimed in claim 1, characterized in that: The system comprises: A disassembly module is used to obtain a target video in real time through a smart access control device, and disassemble the target video frame by frame to obtain a number of continuous single-frame disassembly images; A de-impurity module, used for acquiring a face template image, and performing image de-impurity processing on the disassembled image based on the face template image to obtain a de-impurity image; A first extraction module is used to extract biometric features from the cleaned image to obtain facial biometric features, and perform quality screening on the facial biometric features to obtain screening features; A second extraction module is used to extract facial features from the cleaned image to obtain facial recognition features; An authentication module, used for weightedly fusing the screening feature with the face recognition feature to obtain a weighted feature, performing identity authentication on the target based on the weighted feature, and outputting an identity authentication result; The access control module is used to, if the identity authentication result is that the identity authentication is passed, index the two-ticket database of the corresponding time period based on the identity authentication result and verify the legitimacy of the two-ticket information according to the location of the smart access control device and the identity authentication result; if there are two-ticket data corresponding to the identity authentication result and the two-ticket information is legal, the smart access control device is controlled to be in an open state; if there are no two-ticket data corresponding to the identity authentication result or the identity authentication result is that the identity authentication is failed or the two-ticket information is illegal, the smart access control device is controlled to be in a closed state.

7. A computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the identity authentication control method based on smart access control and two-ticket linkage as described in any one of claims 1 to 5.

8. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by a processor, implements the identity authentication control method based on smart access control and two-ticket linkage as described in any one of claims 1 to 5.

Images