A multimodal identity recognition intelligent access control system and device

Abstract

This invention relates to the field of access control technology, specifically to an intelligent access control system and device with multimodal identity recognition, used to implement the following steps: determining the comprehensive recognition response of each target person based on their identity recognition status under different preset identity recognition methods; clustering all target people and determining the recognition priority value for each target cluster; determining the overall permission index for each target access point; determining the current traffic congestion estimate; and determining the number of current identity recognition methods for each target access point based on the recognition priority value, the overall permission index for each target access point, and the current traffic congestion estimate, thereby achieving intelligent access control. This invention, by analyzing personnel traffic conditions, adjusts the number of current identity recognition methods and improves the rationality of this adjustment, thus improving the rationality of access control.

Description

Technical Field

[0001] This invention relates to the field of access control registration technology, specifically to an intelligent access control system and device with multimodal identity recognition. Background Technology

[0002] Multimodal identity recognition-based intelligent access control systems have received widespread attention in recent years, especially driven by the rapid development of intelligent security and IoT technologies, gradually becoming one of the key technologies for improving security and convenience. With the maturity of technologies such as voice recognition, facial recognition, fingerprint recognition, and iris recognition, multimodal identity recognition systems are increasingly integrating multiple biometric identification methods. This multi-factor authentication approach not only improves system security but also reduces vulnerabilities associated with single-modal technologies.

[0003] In increasingly security-critical enterprise environments, especially those involving personnel access control, identity verification has become a crucial element in ensuring security. Current multimodal identity verification technologies, combining facial recognition, fingerprints, iris scanning, and other biometrics, effectively enhance access security. However, during peak hours, such as commuting times and the end of major events, traditional identity verification processes can face a series of problems due to dense crowds, including cumbersome verification procedures, long verification times per person, queue congestion, and potential impacts on user experience. Therefore, during peak hours, appropriately reducing the number of identity verification methods is essential to improve efficiency. Currently, in situations of congestion, the number of identity verification methods is often adjusted based on human experience, thereby reducing the number of methods used.

[0004] However, when adjusting the number of identity verification methods based on human experience, the following technical problems often arise:

[0005] When adjusting the number of identity recognition methods based on human experience, the results are often influenced by subjective human factors, leading to poor rationality in the adjustment of the number of identity recognition methods, and consequently, poor rationality in access control. Summary of the Invention

[0006] To address the technical problem of poor access control due to the poor rationality of adjusting the number of identity recognition methods, this invention proposes an intelligent access control system and device with multimodal identity recognition.

[0007] In a first aspect, the present invention provides an intelligent access control system for multimodal identity recognition, the system comprising:

[0008] The comprehensive response module is used to determine the comprehensive response of each target person based on their identification performance under different preset identification methods.

[0009] The clustering and determination module is used to cluster all target personnel according to the preset permission level and comprehensive recognition response of all target personnel to obtain target clusters, and to determine the recognition priority value corresponding to each target cluster based on the preset permission level and comprehensive recognition response of the target personnel in each target cluster.

[0010] The overall permission index determination module is used to determine the overall permission index corresponding to each target entrance / exit based on the preset permission value corresponding to each target entrance / exit and the preset permission value corresponding to the entrance / exit connected to it.

[0011] The current traffic congestion prediction module is used to determine the current traffic congestion prediction based on the current preset phase and the historical traffic conditions of people in the previous preset phases.

[0012] The module for determining the number of current identity recognition methods is used to determine the number of current identity recognition methods for each target entrance / exit based on the recognition priority value, the overall permission index corresponding to each target entrance / exit, and the current traffic congestion estimate, in order to realize intelligent entrance / exit control.

[0013] In conjunction with the first aspect above, in one possible implementation, determining the comprehensive recognition response degree for each target person based on their recognition performance under different preset recognition methods includes:

[0014] Based on the identity verification results of each target person when using each preset identity verification method, determine the initial recognition response rate of each target person under each preset identity verification method;

[0015] Based on the total number of verifications for each target person under various preset identity recognition methods, and the initial recognition response rate of each target person under various preset identity recognition methods, the overall recognition response rate for each target person is determined.

[0016] In conjunction with the first aspect above, in one possible implementation, determining the initial recognition response rate of each target person under each preset identity recognition method based on the identity verification status of each target person when using each preset identity recognition method includes:

[0017] Obtain the total number of verification attempts, the number of successful verifications, and the average verification time for each target individual under each preset identity recognition method;

[0018] The average verification time for all target personnel under the same preset identity recognition method is determined as the overall representative verification time corresponding to that preset identity recognition method.

[0019] The ratio of the number of successful verifications for each target person under each preset identity recognition method to the total number of verifications is used to determine the verification smoothness of each target person under each preset identity recognition method.

[0020] The ratio of the average verification time for each target person under each preset identity recognition method to the corresponding overall verification representative time is determined as the recognition time delay for each target person under each preset identity recognition method.

[0021] Based on the verification smoothness and recognition time delay of each target person under each preset identity recognition method, determine the initial recognition response rate of each target person under each preset identity recognition method.

[0022] In conjunction with the first aspect above, in one possible implementation, determining the identification priority value corresponding to each target cluster based on the preset permission level and comprehensive identification response rate of the target personnel within each target cluster includes:

[0023] Based on the average of the preset permission levels of all target personnel within each target cluster and the average of the comprehensive recognition response of all target personnel within each target cluster, the recognition priority value corresponding to each target cluster is determined.

[0024] In conjunction with the first aspect above, in one possible implementation, determining the overall permission index corresponding to each target entrance / exit based on the preset permission value corresponding to each target entrance / exit and the preset permission values ​​corresponding to its connected entrances / exits includes:

[0025] Any target entrance or exit is designated as a marked entrance or exit. From all target entrances or exits, target entrances or exits that are connected to the marked entrance or exit are selected as candidate entrances or exits.

[0026] Filter out the candidate entry / exit with the highest corresponding preset permission value from all candidate entry / exit points, and use it as the reference entry / exit point;

[0027] If the preset permission value corresponding to the reference entry / exit is less than or equal to the preset permission value corresponding to the marked entry / exit, then the preset permission value corresponding to the marked entry / exit is determined as the overall permission index corresponding to the marked entry / exit.

[0028] If the preset permission value corresponding to the reference entrance / exit is greater than the preset permission value corresponding to the marked entrance / exit, then based on the length of the minimum reachable path between the marked entrance / exit and the reference entrance / exit, and the difference between the preset permission value corresponding to the reference entrance / exit and the preset permission value corresponding to the marked entrance / exit, the risk permission supplementary value corresponding to the marked entrance / exit is determined, and the sum of the preset permission value corresponding to the marked entrance / exit and the risk permission supplementary value is determined as the overall permission index corresponding to the marked entrance / exit.

[0029] In conjunction with the first aspect above, in one possible implementation, determining the current traffic congestion estimate based on the historical pedestrian traffic conditions of the current preset phase and previous preset phases includes:

[0030] Select time periods that belong to the same preset stage from all preset historical periods to form a sequence of historical time periods corresponding to that preset stage.

[0031] The current time is defined as the preset stage to which it belongs, and each preset stage before the marked stage is defined as the reference stage.

[0032] The corresponding period is selected from the current period as the reference period to form the reference period corresponding to the reference period;

[0033] Based on the historical personnel traffic data under the marked stage, determine the stage busyness level corresponding to the marked stage;

[0034] Similarly, based on the historical personnel traffic conditions under each reference stage, the corresponding stage busyness level for each reference stage is determined.

[0035] Normalize the difference between the number of people passing through the reference period corresponding to each reference stage and the average number of people passing through the corresponding historical period sequence to obtain the personnel passing contrast for each reference stage.

[0036] Based on the contrast and stage busyness of the personnel corresponding to all reference stages, the personnel corresponding to the marked stage are determined by the change factor.

[0037] Based on the stage busyness level and personnel traffic change factor corresponding to the marked stage, the current traffic busyness estimate is determined.

[0038] In conjunction with the first aspect above, in one possible implementation, determining the stage busyness level corresponding to the marked stage based on the historical personnel passage data under the marked stage includes:

[0039] The ratio of the number of people passing through each historical period in the historical time sequence corresponding to the marked stage to the number of people passing through in its preset historical period is determined as the traffic congestion ratio of the marked stage in each historical period of its corresponding historical time sequence.

[0040] The average of the traffic congestion ratio of the marked stage in all historical time periods of its corresponding historical time period sequence is determined as the stage congestion performance degree corresponding to the marked stage.

[0041] In conjunction with the first aspect above, in one possible implementation, determining the number of current identity recognition methods corresponding to each target entrance / exit based on the recognition priority value, the overall permission index corresponding to each target entrance / exit, and the current traffic congestion estimate includes:

[0042] The identification priority value corresponding to the target cluster to which each target person belongs is determined as the identification priority representative value corresponding to each target person;

[0043] The target personnel waiting to pass through each target entrance / exit at the current moment are identified as the personnel waiting to pass through each target entrance / exit.

[0044] Based on the identification priority representative value of all personnel waiting to pass through each target entrance / exit, as well as the overall permission index and current traffic congestion estimate for each target entrance / exit, determine the number of current identification methods for each target entrance / exit.

[0045] In conjunction with the first aspect above, in one possible implementation, determining the number of current identity recognition methods for each target entrance / exit based on the identification priority representative value corresponding to all personnel waiting to pass through each target entrance / exit, as well as the overall permission index and current traffic congestion estimate for each target entrance / exit, includes:

[0046] The average of the identification priority representative values ​​of all personnel waiting to pass through each target entrance / exit is determined as the current identification priority index for each target entrance / exit.

[0047] Based on the number of preset identity recognition methods, the estimated current traffic congestion, and the current recognition priority and overall permission indicators corresponding to each target entrance / exit, determine the number of current identity recognition methods corresponding to each target entrance / exit.

[0048] Secondly, a multimodal identity recognition intelligent access control device is provided, including a processor and a memory. The processor is used to process instructions stored in the memory to implement the steps of the multimodal identity recognition intelligent access control system.

[0049] Thirdly, the present invention provides the steps of a multimodal identity recognition-based intelligent access control method implemented in a multimodal identity recognition intelligent access control system, comprising:

[0050] Based on the identification performance of different target personnel under different preset identification methods, determine the overall identification response rate for each target personnel;

[0051] Based on the preset permission level and comprehensive recognition response of all target personnel, all target personnel are clustered to obtain target clusters. Based on the preset permission level and comprehensive recognition response of the target personnel in each target cluster, the recognition priority value corresponding to each target cluster is determined.

[0052] Based on the preset permission value corresponding to each target entrance / exit and the preset permission value corresponding to the entrance / exit connected to it, determine the overall permission index corresponding to each target entrance / exit;

[0053] The current traffic congestion level is determined based on the current preset phase and the historical traffic flow data of previous preset phases.

[0054] Based on the identification priority value, the overall permission index corresponding to each target entrance / exit, and the current traffic congestion estimate, the number of current identity recognition methods corresponding to each target entrance / exit is determined to achieve intelligent entrance / exit control.

[0055] Fourthly, a server is provided, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, causing the device to execute the aforementioned intelligent access control method for multimodal identity recognition.

[0056] Fifthly, a computer program product is provided, comprising: computer program code, which, when run on a computer, causes the computer to execute the aforementioned intelligent access control method for multimodal identity recognition.

[0057] Sixthly, a computer-readable storage medium is provided, which stores computer program code that, when executed on a computer, causes the computer to perform the aforementioned intelligent access control method for multimodal identity recognition.

[0058] The present invention has the following beneficial effects:

[0059] This invention discloses a multimodal identity recognition intelligent access control system. By analyzing personnel traffic flow, the identification and verification status and permission levels of different personnel, and the permission status of different access points, it adjusts the number of current identity recognition methods. This solves the technical problem of poor access control caused by poor rationality in adjusting the number of identity recognition methods, and improves the rationality of the adjustment of the number of identity recognition methods, thereby improving the rationality of access control. Specifically, this invention analyzes the identity recognition status of different target personnel under different identity recognition methods, quantifies the comprehensive recognition response degree representing the identification and verification response status of different target personnel, and groups target personnel with similar identification and verification response status and preset permission levels into the same target cluster, thereby quantifying the recognition priority value corresponding to different target clusters. Secondly, it quantifies the overall permission indicators related to access permissions and quantifies the current traffic congestion prediction degree representing the current traffic congestion status. This relatively objectively realizes the adjustment of the number of identity recognition methods corresponding to different target access points, improving the rationality of the adjustment of the number of identity recognition methods to a certain extent, and thus improving the rationality of access control. Attached Figure Description

[0060] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0061] Figure 1 This is a schematic diagram of the structure of an intelligent access control system for multimodal identity recognition according to the present invention;

[0062] Figure 2 This is a flowchart of an intelligent access control method for multimodal identity recognition according to the present invention;

[0063] Figure 3 This is a schematic diagram of the structure of a computer device according to the present invention. Detailed Implementation

[0064] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the technical solution proposed according to the present invention are described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0065] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0066] refer to Figure 1 A schematic diagram of a multimodal identity recognition intelligent access control system according to the present invention is shown. This multimodal identity recognition intelligent access control system includes:

[0067] The comprehensive recognition response module 101 is used to determine the comprehensive recognition response of each target person based on the recognition performance of different target persons under different preset identity recognition methods.

[0068] The target personnel can be individuals who have registered their identity information with the company. For example, the target personnel can be employees within the same company. The preset identity recognition method can be a single identity recognition method from a pre-set multimodal identity recognition system. The preset identity recognition method can be pre-set, and it can be up to 5. For example, the preset identity recognition method can be, but is not limited to: facial recognition, fingerprint recognition, finger vein verification, password verification, and dynamic password verification.

[0069] As an example, determining the overall recognition response for each target person may include the following steps:

[0070] The first step, based on the identity verification results of each target person when using each preset identity verification method, determines the initial recognition response rate of each target person under each preset identity verification method, which may include the following sub-steps:

[0071] The first sub-step is to obtain the total number of verifications, the number of successful verifications, and the average verification time for each target person under each preset identity recognition method.

[0072] For example, taking facial recognition as an example, any target person can be identified as a labeled person, and the total number of times the labeled person has used facial recognition for verification in the past can be recorded as the total number of verifications by the labeled person under facial recognition; the total number of times the labeled person has successfully verified using facial recognition in the past can be recorded as the number of successful verifications by the labeled person under facial recognition; the time taken for the labeled person to complete each verification using facial recognition in the past can be recorded as the labeling time, and the average of all labeling times can be recorded as the average verification time of the labeled person under facial recognition.

[0073] It should be noted that the verification time may vary for different people even under the same identification method. For example, for facial recognition, changes in hairstyle, whether a person is wearing a mask, or whether they are wearing a hat may all affect the verification time.

[0074] The second sub-step is to determine the average verification time of all target personnel under the same preset identity recognition method as the overall representative verification time corresponding to that preset identity recognition method.

[0075] For example, the formula for determining the overall verification time corresponding to a preset identity recognition method can be:

[0076]

[0077] Among them, A j This represents the total verification time for the j-th preset identity recognition method. j is the type number of the preset identity recognition method. N is the number of target personnel. i is the serial number of the target personnel. A ij It is the average verification time for the i-th target person under the j-th preset identity recognition method.

[0078] It should be noted that A j It can characterize the average usage time of the j-th preset identity recognition method. The larger the value, the slower the verification and recognition speed of different people using the j-th preset identity recognition method.

[0079] The third sub-step is to determine the verification smoothness of each target person under each preset identity recognition method by the ratio of the number of successful verifications to the total number of verifications.

[0080] For example, the formula for determining the verification smoothness of different target personnel under different preset identity recognition methods can be:

[0081]

[0082] Among them, B ij This represents the verification success rate of the i-th target person under the j-th preset identity verification method. i is the target person's serial number. j is the preset identity verification method number. ij1 b represents the number of times the i-th target person has been successfully verified using the j-th preset identity verification method. ij2 It represents the total number of verifications for the i-th target person under the j-th preset identity recognition method.

[0083] It should be noted that when B ij The larger the value, the smoother the verification and recognition of the i-th target person under the j-th preset identity recognition method.

[0084] The fourth sub-step is to determine the recognition time delay for each target person under each preset identity recognition method by comparing the average verification time of each target person under each preset identity recognition method with the corresponding overall verification representative time.

[0085] For example, the formula for determining the recognition time delay of different target personnel under different preset identity recognition methods can be:

[0086]

[0087] Among them, C ij This represents the identification delay time for the i-th target person under the j-th preset identification method. i is the target person's index. j is the preset identification method number. A ij It represents the average verification time for the i-th target person under the j-th preset identity recognition method. j It represents the overall verification time corresponding to the j-th preset identity recognition method.

[0088] It should be noted that A j This can characterize the average usage time of the j-th preset identity recognition method. A ij This can characterize the average usage time of the i-th target person under the j-th preset identity recognition method. Therefore, when C ij A larger value usually indicates that the average usage time of the i-th target person under the j-th preset identity recognition method is relatively longer.

[0089] The fifth sub-step involves determining the initial recognition response rate of each target person under each preset identity recognition method based on the verification smoothness and recognition time delay of each target person under each preset identity recognition method.

[0090] For example, the formula for determining the initial recognition response rate of different target individuals under different preset identity recognition methods can be:

[0091] D ij =norm(exp(-B ij )×C ij );

[0092] Among them, D ij This represents the initial recognition response rate of the i-th target person under the j-th preset identification method. i is the target person's index. j is the type of preset identification method. `norm()` is the normalization function. `exp()` is an exponential function with the natural constant as its base. B ij C represents the smoothness of verification for the i-th target person under the j-th preset identity recognition method. ij It is the recognition time delay of the i-th target person under the j-th preset identity recognition method.

[0093] It should be noted that when B ij The smaller the value, the less smooth the verification and identification of the i-th target person under the j-th preset identity recognition method is, which to some extent increases the verification time cost. When C ij A larger value for D generally indicates a longer average usage time for the i-th target person under the j-th preset identification method. Therefore, when D... ij The larger the value, the higher the time cost for the i-th target person to be verified and identified under the j-th preset identity recognition method.

[0094] The second step is to determine the overall recognition response rate for each target person based on the total number of verifications for each target person under various preset identity recognition methods and the initial recognition response rate for each target person under various preset identity recognition methods.

[0095] For example, the formula for determining the comprehensive response score of the target person can be:

[0096]

[0097] Among them, E i This represents the overall recognition response rate for the i-th target person. i is the target person's index. `norm()` is the normalization function. `n` is the number of preset identity recognition methods. `j` is the type number of the preset identity recognition method. ij2 n is the total number of verification attempts for the i-th target person under the j-th preset identity verification method. i D is the cumulative value of the total number of verifications for the i-th target person under all preset identity recognition methods. ij It is the initial recognition response rate of the i-th target person under the j-th preset identity recognition method.

[0098] It should be noted that when D ij A larger value generally indicates a higher time cost for verifying the i-th target person using the j-th preset identity recognition method. The larger the value, the more likely the j-th preset identity recognition method is to be the preset identity recognition method commonly used by the i-th target person, and the more valuable it is for reference. Can be used as D ij The weight of E. Therefore, when E i The larger the value, the higher the time cost for verifying the i-th target person under different preset identity recognition methods. This usually indicates that the priority of verifying the i-th target person should be appropriately increased during congestion.

[0099] The clustering and determination module 102 is used to cluster all target personnel according to the preset permission level and recognition comprehensive response degree of all target personnel to obtain target clusters, and to determine the recognition priority value corresponding to each target cluster based on the preset permission level and recognition comprehensive response degree of the target personnel in each target cluster.

[0100] The preset permission level can be a pre-defined permission level for personnel, with a higher value indicating higher permissions. For example, permission levels can be set according to the target personnel's job level within the company; the higher the job level, the higher the preset permission level. For instance, a leader's permission level can be higher than that of a regular employee.

[0101] As an example, the clustering and determination module 102 can specifically implement the following steps:

[0102] The first step is to cluster all target personnel according to their preset permission levels and overall recognition response, thus obtaining target clusters.

[0103] For example, based on the preset permission level and comprehensive recognition response of all target personnel, the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm is used to cluster all target personnel, and the resulting clusters are recorded as target clusters.

[0104] The second step is to determine the recognition priority value for each target cluster based on the average of the preset permission levels of all target personnel within each target cluster and the average of the comprehensive recognition response of all target personnel within each target cluster.

[0105] For example, the formula for determining the recognition priority value corresponding to the target cluster can be:

[0106] F a =norm(E a ×Q a );

[0107] Among them, F a This is the recognition priority value corresponding to the a-th target cluster. 'a' is the index of the target cluster. 'norm()' is the normalization function. E a Q is the average of the comprehensive recognition response scores for all target personnel within the a-th target cluster. a It is the average of the preset permission levels corresponding to all target personnel within the a-th target cluster.

[0108] It should be noted that when E aA larger Q value generally indicates a higher time cost for verifying and identifying target personnel within the a-th target cluster under different preset identity recognition methods. This often suggests that the priority for identifying and verifying target personnel within the a-th target cluster should be appropriately increased during congestion. a A larger value for F generally indicates higher privileges for personnel within the a-th target cluster, suggesting that the priority for identifying and verifying personnel within the a-th target cluster should be appropriately increased during congestion. Therefore, when F... a A higher value usually indicates that the priority of identifying and verifying target personnel within the a-th target cluster should be appropriately increased during periods of congestion.

[0109] The overall permission index determination module 103 is used to determine the overall permission index corresponding to each target entrance / exit based on the preset permission value corresponding to each target entrance / exit and the preset permission value corresponding to the entrance / exit connected to it.

[0110] The preset permission value can be a pre-set permission value. The higher the preset permission value corresponding to the target entrance / exit, the more important the target entrance / exit is, and the higher the security level needs to be set. For example, the access permission of the entrance / exit of the company's internal document room is often higher than that of the entrance / exit of the company's living room, so the preset permission value of the entrance / exit of the company's internal document room can be set to be higher than that of the entrance / exit of the company's living room.

[0111] As an example, determining the overall access control metrics for each target entry / exit point may include the following steps:

[0112] The first step is to designate any target entrance / exit as a marked entrance / exit, and then select from all target entrances / exits that are connected to the marked entrance / exit as candidate entrances / exits.

[0113] Among them, the marked entrances and exits can be the entrances and exits that the candidate entrances and exits pass through.

[0114] The second step is to filter out the candidate entry / exit with the highest preset permission value from all candidate entry / exit points, and use it as the reference entry / exit point.

[0115] Third, if the preset permission value corresponding to the above-mentioned reference entry / exit is less than or equal to the preset permission value corresponding to the above-mentioned marked entry / exit, then the preset permission value corresponding to the above-mentioned marked entry / exit is determined as the overall permission index corresponding to the above-mentioned marked entry / exit.

[0116] Fourth, if the preset permission value corresponding to the reference entrance / exit is greater than the preset permission value corresponding to the marked entrance / exit, then based on the length of the minimum reachable path between the marked entrance / exit and the reference entrance / exit, and the difference between the preset permission value corresponding to the reference entrance / exit and the preset permission value corresponding to the marked entrance / exit, determine the risk permission supplementary value corresponding to the marked entrance / exit, and determine the sum of the preset permission value corresponding to the marked entrance / exit and the risk permission supplementary value as the overall permission index corresponding to the marked entrance / exit.

[0117] The method for obtaining the length of the minimum reachable path between the marked entrance / exit and the reference entrance / exit can be as follows: each path formed between the marked entrance / exit and the reference entrance / exit is identified as a marked path, and the length of each marked path is recorded as the marked length. The minimum value among all marked lengths is determined as the length of the minimum reachable path between the marked entrance / exit and the reference entrance / exit.

[0118] For example, the formula for determining the supplementary risk permission value corresponding to the marked entry / exit point can be:

[0119] H = exp(-S) × (q1 - q0);

[0120] Where H is the supplementary risk permission value corresponding to the marked entry / exit. exp() is an exponential function with the natural constant as its base. S is the length of the minimum reachable path between the marked entry / exit and the reference entry / exit. q1 is the preset permission value corresponding to the reference entry / exit. q0 is the preset permission value corresponding to the marked entry / exit.

[0121] It should be noted that a smaller S usually indicates that the marked entry / exit is closer to a reference entry / exit with higher privileges. A larger q1-q0 usually indicates that the reference entry / exit connected to the marked entry / exit has relatively higher privileges. Therefore, a larger H usually indicates that because the marked entry / exit is closer to a reference entry / exit with higher privileges, the privileges of the marked entry / exit should be appropriately increased.

[0122] The current traffic congestion prediction module 104 is used to determine the current traffic congestion prediction based on the current preset stage and the historical traffic conditions of people in the previous preset stages.

[0123] The preset stage can be a stage within a pre-set time period, with a corresponding duration of half an hour. The time period can be a day. For example, if the duration of a preset stage is half an hour, then a day of 24 hours can be divided into 48 stages, such as 8:30 to 9:00 as a preset stage.

[0124] As an example, determining the current traffic congestion estimate may include the following steps:

[0125] The first step is to select time periods that belong to the same preset stage from all preset historical periods, and then construct the historical time period sequence corresponding to that preset stage.

[0126] The preset historical period can be a pre-set time period that has already passed. The last preset historical period can be yesterday. The number of preset historical periods can be pre-set, and can be 30.

[0127] For example, if the preset period is from 8:30 to 9:00, then all time periods falling between 8:30 and 9:00 can be selected from the past 30 days to form a historical time period sequence corresponding to the preset period. In this case, the number of time periods in the historical time period sequence can be 30. The time periods in the historical time period sequence can be referred to as historical time periods.

[0128] The second step is to determine the current time to be the preset stage as the marking stage, and to determine each preset stage before the marking stage as the reference stage.

[0129] It should be noted that the marking phase can be the current phase of the day, while the reference phase can be a phase that has already passed on the current day.

[0130] The third step is to select the corresponding period from the current preset period as the reference period, thus forming the reference period corresponding to the reference period.

[0131] The preset current period can be a pre-defined time period to which the current moment belongs, which can be the current day.

[0132] The fourth step, based on the historical personnel traffic data under the aforementioned marked stages, determines the stage busyness level corresponding to the marked stages, which may include the following sub-steps:

[0133] The first sub-step is to determine the ratio of the number of people passing through each historical period in the historical time sequence corresponding to the above-mentioned marked stage to the number of people passing through in the preset historical period to the above-mentioned marked stage as the traffic congestion ratio of the above-mentioned marked stage in each historical time period in its corresponding historical time sequence.

[0134] The second sub-step is to determine the average of the traffic congestion ratio of the marked stage in all historical time periods in its corresponding historical time period sequence as the stage congestion performance degree corresponding to the marked stage.

[0135] For example, the formula for determining the busy performance level of the corresponding marking stage can be:

[0136]

[0137] Where G represents the busy performance level of the marking phase. M represents the number of historical time periods in the historical time period sequence corresponding to the marking phase. m represents the sequence number of the historical time period in the historical time period sequence corresponding to the marking phase. d m1 This refers to the number of people passing through the m-th historical time period in the historical time period sequence corresponding to the marking phase; that is, the total number of people passing through all target entrances / exits in the m-th historical time period. Whether a person enters or exits a target entrance / exit, it represents that the person has passed through that entrance / exit. m2 It is the number of people passing through the m-th historical period in the historical time sequence corresponding to the stage, which is the total number of people passing through all target entrances and exits in the m-th historical period within the preset historical cycle. It is the ratio of busy traffic to pedestrian traffic in the m-th historical period of the corresponding historical period sequence of the marked stage.

[0138] It should be noted that when G is larger, it often indicates that the number of people passing through during the marking phase each day is relatively larger, which often indicates that the marking phase is more likely to be a peak period.

[0139] Fifth, similarly, determine the busy performance level of each reference stage based on the historical personnel traffic conditions under each reference stage.

[0140] It should be noted that the method for obtaining the busy performance level of the reference stage can be the same as the method for obtaining the busy performance level of the marked stage, and will not be repeated here.

[0141] The sixth step is to normalize the difference between the number of people passing through the reference period corresponding to each reference stage and the average number of people passing through the corresponding historical period sequence, so as to obtain the personnel passing contrast corresponding to each reference stage.

[0142] For example, the formula for determining the personnel corresponding to the reference stage based on the contrast ratio can be:

[0143] L p =norm(d p1 -d p );

[0144] Among them, L p This refers to the personnel in the p-th reference stage who passed the comparison. p is the reference stage number. `norm()` is the normalization function. d p1 d represents the number of people passing through during the reference time period corresponding to the p-th reference stage, which is the total number of people passing through all target entrances and exits during the reference time period corresponding to the p-th reference stage. pIt is the average number of people passing through in the historical time period sequence corresponding to the p-th reference stage, which is the average number of people passing through all historical time periods in the historical time period sequence corresponding to the p-th reference stage for all target entrances and exits.

[0145] It should be noted that when L p The larger the value, the higher the number of people passing through in the p-th reference stage on that day is compared to the number of people passing through in the p-th reference stage in historical time, which often indicates that the number of people passing through on that day is more likely to show an increasing trend.

[0146] Step 7: Based on the contrast and busy performance of personnel in all reference stages, determine the personnel change factor corresponding to the marked stages mentioned above.

[0147] For example, the formula for determining the personnel corresponding to the marking stage using the change factor can be:

[0148]

[0149] Where Y is the change factor for the personnel corresponding to the marking stage. y is the number of reference stages. p is the index of the reference stage. norm() is the normalization function. G p This represents the stage busyness performance corresponding to the p-th reference stage. L p The personnel corresponding to the p-th reference stage are compared through contrast.

[0150] It should be noted that when G p A larger value for L usually indicates a relatively larger number of people passing through in the p-th reference phase each day, suggesting that the p-th reference phase is more likely to be a peak period. p A larger value (norm(G)) generally indicates that the number of people passing through in the p-th reference phase on that day is relatively higher than the number of people passing through in the p-th reference phase over historical periods, and that the number of people passing through on that day is more likely to show an increasing trend. p )×L p To a certain extent, Y represents the degree of personnel growth during the p-th reference phase of a given day. Therefore, a larger Y usually indicates a greater degree of personnel growth on that day.

[0151] Step 8: Based on the stage busyness performance and personnel traffic change factor corresponding to the above-mentioned marked stages, determine the current traffic busyness estimate.

[0152] For example, the formula for determining the current traffic congestion estimate can be:

[0153]

[0154] in, G represents the current traffic congestion estimate. Y represents the traffic congestion performance level corresponding to the marked phase.

[0155] It should be noted that a larger Y value generally indicates a greater increase in personnel on that day. A larger G value generally indicates a relatively larger number of people passing through during the marking phase each day, suggesting that the marking phase is more likely to be a peak period, and therefore, the current moment is more likely to be a peak period. Therefore, when The larger the value, the more likely the current stage is to be a peak period, and the more necessary it is to appropriately reduce the number of identity verification methods in order to improve the overall efficiency of verification and recognition.

[0156] The current identity recognition method determination module 105 is used to determine the current identity recognition method for each target entrance / exit based on the recognition priority value, the overall permission index corresponding to each target entrance / exit, and the current traffic congestion estimate, so as to realize intelligent entrance / exit control.

[0157] As an example, determining the number of current identification methods corresponding to each target entrance / exit may include the following steps:

[0158] The first step is to determine the identification priority value corresponding to the target cluster to which each target person belongs, and then determine the identification priority representative value corresponding to each target person.

[0159] The second step is to identify the target personnel waiting to pass through each target entrance / exit at the current moment.

[0160] The third step, based on the identification priority representative value corresponding to all personnel waiting to pass through each target entrance / exit, as well as the overall permission index and current traffic congestion estimate for each target entrance / exit, determines the number of current identification methods for each target entrance / exit, which may include the following sub-steps:

[0161] The first sub-step is to determine the average of the identification priority representative values ​​of all personnel waiting to pass through each target entrance / exit as the current identification priority index for each target entrance / exit.

[0162] The second sub-step involves determining the number of current identity recognition methods for each target entrance / exit based on the number of preset identity recognition methods, the estimated current traffic congestion level, and the current recognition priority and overall permission indicators for each target entrance / exit.

[0163] For example, the formula for determining the number of current identity recognition methods corresponding to a target entrance / exit can be:

[0164]

[0165] Among them, R r It represents the number of current identity verification methods corresponding to the r-th target entrance / exit. r is the sequence number of the target entrance / exit. It is the floor function. `norm()` is the normalization function. W r This is the overall permission index corresponding to the r-th target entry / exit. exp() is an exponential function with the natural constant as its base. F r It is the current identification priority index corresponding to the r-th target entrance / exit, which is the average of the identification priority representative values ​​of all personnel waiting to pass through the r-th target entrance / exit. R represents the current traffic congestion estimate. R is the number of preset identification methods.

[0166] It should be noted that when F r A larger value for W generally indicates that during congestion, the priority of identity verification at the r-th target entrance / exit should be increased, and the number of identification methods at the r-th target entrance / exit should be reduced to improve overall verification efficiency. r A larger value often indicates that the r-th target entrance / exit is closer to a higher-privilege entrance / exit, and that the r-th target entrance / exit itself likely has higher privileges; this often means that the r-th target entrance / exit needs a higher security level, and requires more identity verification methods. A larger value for R often indicates that the current stage is more likely to be a peak period, suggesting that it's necessary to appropriately reduce the number of identity verification methods at that stage to improve overall verification efficiency. Therefore, R... r It can represent the number of identity recognition methods required at the r-th target entrance / exit within the current phase.

[0167] Optionally, based on the number of current identity recognition methods corresponding to the target entrance / exit, the intelligent entrance / exit control can include: arranging all preset identity recognition methods in ascending order according to the overall verification representative time, obtaining a preset identity recognition method sequence, and recording the current number of identity recognition methods as the target number, recording the first preset number of preset identity recognition methods in the preset identity recognition method sequence as the target identity recognition method, so that the target entrance / exit can use the target identity recognition method for verification and recognition at the current moment, instead of continuing to use other identity recognition methods for verification and recognition, thereby improving the overall verification and recognition efficiency while ensuring the security of verification and recognition as much as possible.

[0168] Based on the same inventive concept as the above embodiments, the present invention provides a multimodal identity recognition intelligent access control device, including a processor and a memory. The processor is used to process instructions stored in the memory to implement the steps of the above-described multimodal identity recognition intelligent access control system.

[0169] refer to Figure 2 Based on the same inventive concept as the above-described method embodiments, this invention provides a multimodal identity recognition intelligent access control method, comprising the following steps:

[0170] Step S1: Determine the overall recognition response rate for each target person based on their recognition performance under different preset recognition methods.

[0171] Step S2: Based on the preset permission level and comprehensive recognition response of all target personnel, cluster all target personnel to obtain target clusters, and determine the recognition priority value corresponding to each target cluster based on the preset permission level and comprehensive recognition response of the target personnel in each target cluster.

[0172] Step S3: Determine the overall permission index for each target entrance / exit based on the preset permission value corresponding to each target entrance / exit and the preset permission values ​​corresponding to the entrances / exits connected to it.

[0173] Step S4: Determine the current traffic congestion level based on the current preset phase and the historical traffic conditions of previous preset phases.

[0174] Step S5: Based on the identification priority value, the overall permission index corresponding to each target entrance / exit, and the current traffic congestion estimate, determine the number of current identity recognition methods corresponding to each target entrance / exit, so as to realize intelligent entrance / exit control.

[0175] Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. For example, as shown... Figure 3 As shown, the computer device 300 includes: a memory 301, a processor 302, and a computer program 303 stored in the memory 301 and running on the processor 302. When the processor 302 executes the computer program 303, the computer device can execute the aforementioned intelligent access control method for multimodal identity recognition.

[0176] Based on the same inventive concept as the above-described method embodiments, the present invention provides a server, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, causing the device to execute the aforementioned intelligent access control method for multimodal identity recognition.

[0177] Based on the same inventive concept as the above-described method embodiments, the present invention provides a computer program product comprising: computer program code, which, when executed on a computer, causes the computer to perform the above-described intelligent access control method for multimodal identity recognition.

[0178] Based on the same inventive concept as the above-described method embodiments, the present invention provides a computer-readable storage medium storing computer program code, which, when executed on a computer, causes the computer to perform the above-described intelligent access control method for multimodal identity recognition.

[0179] In summary, this invention analyzes the identity recognition performance of different target personnel under different recognition methods, quantifies the comprehensive recognition response rate (CRSR) characterizing the recognition and verification response of different target personnel, and groups target personnel with similar recognition and verification response rates and preset permission levels into the same target cluster, thereby quantifying the recognition priority value corresponding to different target clusters. Secondly, it quantifies the overall permission indicators related to entrance and exit permissions and quantifies the current traffic congestion prediction rate (CHD), thus relatively objectively adjusting the number of identity recognition methods corresponding to different target entrances and exits. This improves the rationality of adjusting the number of identity recognition methods to a certain extent, thereby enhancing the rationality of entrance and exit control.

[0180] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A multimodal identity recognition intelligent access control system, characterized in that, The system includes: The comprehensive response module is used to determine the comprehensive response of each target person based on their identification performance under different preset identification methods. The clustering and determination module is used to cluster all target personnel according to the preset permission level and comprehensive recognition response of all target personnel to obtain target clusters, and to determine the recognition priority value corresponding to each target cluster based on the preset permission level and comprehensive recognition response of the target personnel in each target cluster. The overall permission index determination module is used to determine the overall permission index corresponding to each target entrance / exit based on the preset permission value corresponding to each target entrance / exit and the preset permission value corresponding to the entrance / exit connected to it. The current traffic congestion prediction module is used to determine the current traffic congestion prediction based on the current preset phase and the historical traffic conditions of people in the previous preset phases. The current identity recognition method determination module is used to determine the current identity recognition method for each target entrance / exit based on the recognition priority value, the overall permission index corresponding to each target entrance / exit, and the current traffic congestion estimate, so as to realize intelligent entrance / exit control. The step of determining the overall recognition response rate for each target person based on their identity recognition performance under different preset identity recognition methods includes: determining the initial recognition response rate for each target person under each preset identity recognition method based on their identity verification performance when using each preset identity recognition method; and determining the overall recognition response rate for each target person based on the total number of verifications for each target person under various preset identity recognition methods and the initial recognition response rate for each target person under various preset identity recognition methods, which are obtained in advance.

2. The intelligent access control system for multimodal identity recognition according to claim 1, characterized in that, The step of determining the initial recognition response rate of each target person under each preset identity recognition method based on the identity verification status of each target person when using each preset identity recognition method includes: Obtain the total number of verification attempts, the number of successful verifications, and the average verification time for each target individual under each preset identity recognition method; The average verification time for all target personnel under the same preset identity recognition method is determined as the overall representative verification time corresponding to that preset identity recognition method. The ratio of the number of successful verifications for each target person under each preset identity recognition method to the total number of verifications is used to determine the verification smoothness of each target person under each preset identity recognition method. The ratio of the average verification time for each target person under each preset identity recognition method to the corresponding overall verification representative time is determined as the recognition time delay for each target person under each preset identity recognition method. Based on the verification smoothness and recognition time delay of each target person under each preset identity recognition method, determine the initial recognition response rate of each target person under each preset identity recognition method.

3. The intelligent access control system for multimodal identity recognition according to claim 1, characterized in that, The determination of the identification priority value for each target cluster based on the preset permission level and comprehensive identification response rate of the target personnel within each target cluster includes: Based on the average of the preset permission levels of all target personnel within each target cluster and the average of the comprehensive recognition response of all target personnel within each target cluster, the recognition priority value corresponding to each target cluster is determined.

4. The intelligent access control system for multimodal identity recognition according to claim 1, characterized in that, The step of determining the overall permission index for each target entrance / exit based on the preset permission value corresponding to each target entrance / exit and the preset permission values ​​corresponding to its connected entrances / exits includes: Any target entrance or exit is designated as a marked entrance or exit. From all target entrances or exits, target entrances or exits that are connected to the marked entrance or exit are selected as candidate entrances or exits. Filter out the candidate entry / exit with the highest corresponding preset permission value from all candidate entry / exit points, and use it as the reference entry / exit point; If the preset permission value corresponding to the reference entry / exit is less than or equal to the preset permission value corresponding to the marked entry / exit, then the preset permission value corresponding to the marked entry / exit is determined as the overall permission index corresponding to the marked entry / exit. If the preset permission value corresponding to the reference entrance / exit is greater than the preset permission value corresponding to the marked entrance / exit, then based on the length of the minimum reachable path between the marked entrance / exit and the reference entrance / exit, and the difference between the preset permission value corresponding to the reference entrance / exit and the preset permission value corresponding to the marked entrance / exit, the risk permission supplementary value corresponding to the marked entrance / exit is determined, and the sum of the preset permission value corresponding to the marked entrance / exit and the risk permission supplementary value is determined as the overall permission index corresponding to the marked entrance / exit.

5. The intelligent access control system for multimodal identity recognition according to claim 1, characterized in that, The step of determining the current traffic congestion estimate based on the current preset phase and the historical traffic conditions of previous preset phases includes: Select time periods that belong to the same preset stage from all preset historical periods to form a sequence of historical time periods corresponding to that preset stage. The current time is defined as the preset stage to which it belongs, and each preset stage before the marked stage is defined as the reference stage. The corresponding period is selected from the current period as the reference period to form the reference period corresponding to the reference period; Based on the historical personnel traffic data under the marked stage, determine the stage busyness level corresponding to the marked stage; Similarly, based on the historical personnel traffic conditions under each reference stage, the corresponding stage busyness level for each reference stage is determined. Normalize the difference between the number of people passing through the reference period corresponding to each reference stage and the average number of people passing through the corresponding historical period sequence to obtain the personnel passing contrast for each reference stage. Based on the contrast and stage busyness of the personnel corresponding to all reference stages, the personnel corresponding to the marked stage are determined by the change factor. Based on the stage busyness level and personnel traffic change factor corresponding to the marked stage, the current traffic busyness estimate is determined.

6. The intelligent access control system for multimodal identity recognition according to claim 5, characterized in that, The step of determining the busyness level of the marked stage based on the historical personnel traffic data under the marked stage includes: The ratio of the number of people passing through each historical period in the historical time sequence corresponding to the marked stage to the number of people passing through in its preset historical period is determined as the traffic congestion ratio of the marked stage in each historical period of its corresponding historical time sequence. The average of the traffic congestion ratio of the marked stage in all historical time periods of its corresponding historical time period sequence is determined as the stage congestion performance degree corresponding to the marked stage.

7. The intelligent access control system for multimodal identity recognition according to claim 1, characterized in that, The step of determining the number of current identity recognition methods for each target entrance / exit based on the recognition priority value, the overall permission index corresponding to each target entrance / exit, and the current traffic congestion estimate includes: The identification priority value corresponding to the target cluster to which each target person belongs is determined as the identification priority representative value corresponding to each target person; The target personnel waiting to pass through each target entrance / exit at the current moment are identified as the personnel waiting to pass through each target entrance / exit. Based on the identification priority representative value of all personnel waiting to pass through each target entrance / exit, as well as the overall permission index and current traffic congestion estimate for each target entrance / exit, determine the number of current identification methods for each target entrance / exit.

8. The intelligent access control system for multimodal identity recognition according to claim 7, characterized in that, The method of determining the number of current identity recognition methods for each target entrance / exit based on the identification priority representative value corresponding to all personnel waiting to pass through each target entrance / exit, the overall permission index corresponding to each target entrance / exit, and the current traffic congestion estimate includes: The average of the identification priority representative values ​​of all personnel waiting to pass through each target entrance / exit is determined as the current identification priority index for each target entrance / exit. Based on the number of preset identity recognition methods, the estimated current traffic congestion, and the current recognition priority and overall permission indicators corresponding to each target entrance / exit, determine the number of current identity recognition methods corresponding to each target entrance / exit.

9. A multimodal identity recognition intelligent access control device, characterized in that, It includes a processor and a memory, the processor being used to process instructions stored in the memory to implement the steps of the intelligent access control system for multimodal identity recognition as described in any one of claims 1-8.

Images