CN108229260B

CN108229260B - Identity information verification method and system

Info

Publication number: CN108229260B
Application number: CN201611197275.4A
Authority: CN
Inventors: 段广卿
Original assignee: Hangzhou Hikvision System Technology Co Ltd
Current assignee: Hangzhou Hikvision System Technology Co Ltd
Priority date: 2016-12-21
Filing date: 2016-12-21
Publication date: 2020-12-29
Anticipated expiration: 2036-12-21
Also published as: CN108229260A

Abstract

The embodiment of the invention discloses an identity information verification method and an identity information verification system, relates to a face recognition technology, and can improve identity information verification efficiency. The method comprises the following steps: receiving an identity information verification request containing a face image; extracting a face image, and acquiring one or more distributed operation servers meeting the preset load balancing strategy; analyzing and splitting the extracted multiple face images according to the acquired distributed operation server, and correspondingly distributing the face images to the acquired distributed operation server; and receiving comparison result information reported by the acquired distributed operation server, wherein the comparison result information is obtained by the distributed operation server through similarity value calculation of the extracted features in the received face images and the features in a pre-stored feature library, and comprises similarity values and identity information mapped by the similarity values, and each face image corresponds to one piece of comparison result information. The invention is suitable for large-scale identity information verification based on face recognition.

Description

Identity information verification method and system

Technical Field

The invention relates to a face recognition technology, in particular to an identity information verification method and system.

Background

The face recognition technology has been widely applied after years of development, for example, to traffic monitoring, security monitoring, railway and aviation ticket verification, evasion tracking, identity information verification, security inspection, and the like, and a large number of recognition algorithms have been developed. The face recognition technology is a technology for recognizing face features of a person, wherein an input face image is recognized, under the condition that the face exists, the contour (such as eyes, a nose and a mouth) of the face is detected through a specific algorithm according to the face image, implied features are further extracted according to the detected face contour information, the extracted features are compared with a pre-stored feature library (comprising one or more face features) according to a preset feature matching algorithm to obtain one or more similarity values, and accordingly one or more identity information corresponding to the face image is determined according to the similarity values. For example, taking railway security inspection monitoring as an example, by collecting facial images of passengers, extracting features of the collected facial images, and then performing feature matching on the extracted features one by one with a preset feature library by using a feature matching algorithm, the identity information of the passengers is identified, and is compared with the ticket service held by the passengers, so that whether the passengers are consistent with the tickets or not can be determined. Meanwhile, through the feature matching algorithm, suspects or suspicious persons can be identified, so that the safety and the effectiveness of monitoring are effectively improved.

However, in the existing identity information verification method, when feature matching is performed, extracted features are sequentially matched with a preset feature library, so that the comparison speed of feature matching is low, and particularly, when identity information verification is performed on a large-scale face image, the response time of feature matching is long, so that the efficiency of identity information verification is low, and the large-scale (millions and billions) capacity feature matching cannot be met.

Disclosure of Invention

In view of this, embodiments of the present invention provide an identity information verification method and system, which can effectively improve identity information verification efficiency, so as to solve the problems of long response time of feature matching and low identity information verification efficiency caused by that extracted features need to be sequentially matched with a preset feature library in the existing identity information verification method.

In a first aspect, an embodiment of the present invention provides an identity information verification method, including:

receiving an identity information verification request, wherein the identity information verification request comprises a face image;

extracting the face image contained in the identity information verification request, and acquiring one or more distributed operation servers meeting the load balancing strategy according to the extracted face image and a preset load balancing strategy;

splitting the extracted multiple face images according to the acquired distributed operation server, and correspondingly distributing the face images to the acquired distributed operation server;

and receiving comparison result information reported by the acquired distributed operation server, wherein the comparison result information is obtained by the distributed operation server through similarity value calculation of the extracted features in the received face images and the features in a pre-stored feature library, the comparison result information comprises similarity values and identity information mapped by the similarity values, and each face image corresponds to one piece of comparison result information.

With reference to the first aspect, in a first implementation manner of the first aspect, the method further includes:

acquiring and obtaining a face image, and encapsulating the obtained face image in the identity information verification request; and the number of the first and second groups,

and displaying the received comparison result information.

With reference to the first implementation manner of the first aspect, in a second implementation manner of the first aspect, the method further includes:

and storing the comparison result information.

With reference to the second implementation manner of the first aspect, in a third implementation manner of the first aspect, before the splitting the extracted multiple face images according to the acquired distributed computation server, the method further includes:

and performing quality evaluation on the extracted face image, and correcting the face image which does not pass the quality evaluation according to a preset face image correction algorithm.

With reference to the second implementation manner of the first aspect, in a fourth implementation manner of the first aspect, the constructing the pre-stored feature library includes:

receiving a modeling task containing a modeling face image, extracting the features of the modeling face image, and constructing a mapping relation feature library of the extracted features and the identity information according to the identity information corresponding to the modeling face image.

With reference to the first aspect, in a fifth implementation manner of the first aspect, the obtaining, according to the extracted face image and a preset load balancing policy, one or more distributed operation servers that satisfy the load balancing policy includes:

and respectively loading the extracted face images to the distributed operation servers according to the sequence of the load values of the distributed operation servers from small to large.

With reference to the fifth implementation manner of the first aspect, in a sixth implementation manner of the first aspect, the loading the extracted face images to the distributed computation servers in the order from small to large according to the load values of the distributed computation servers includes:

selecting a distributed operation server with the minimum load value to load a part of face images in the extracted face images, so that the load value of the distributed operation server with the minimum load value reaches a preset load threshold value;

and for the distributed operation servers except the distributed operation server with the smallest load value, sequentially loading the rest face images to the distributed operation servers according to the sequence of the load values of the distributed operation servers from small to large, so that all or part of the load values of the rest distributed operation servers reach a preset load threshold value until all the face images are loaded.

With reference to the first aspect, in a seventh implementation manner of the first aspect, the obtaining, according to the extracted face image and a preset load balancing policy, one or more distributed operation servers that satisfy the load balancing policy includes:

acquiring the total load value of all distributed operation servers with communication connection after the extracted face image is loaded;

calculating the quotient of the total load value and the number of the distributed operation servers to obtain an average load value;

and sequentially loading the rest face images of the extracted face images from the distributed operation server with the load value smaller than the average load value according to the sequence of the load value of the distributed operation server from small to large so as to enable the load of the loaded distributed operation server to reach the average load.

With reference to the first aspect, the first or second implementation manner of the first aspect, in an eighth implementation manner of the first aspect, the calculating, by extracting features in the received face image and performing similarity calculation with features in a feature library stored in advance, the comparison result information includes:

traversing each received face image, and extracting features contained in the face image;

calculating similarity values of the extracted features and the features in a pre-stored feature library;

sorting according to the sequence of similarity values from high to low;

selecting similarity values of n bits before sorting, and acquiring the features of the selected similarity values in the feature library;

and inquiring the features in the feature library according to the obtained selected similarity value corresponding to the features in the feature library to obtain the identity information mapped by the obtained features respectively, and packaging the similarity value and the identity information mapped by the similarity value in the comparison result information.

In a second aspect, an embodiment of the present invention provides an identity information verification system, including: a query display subsystem, a service processing subsystem and a comparison subsystem, wherein,

the query display subsystem is used for packaging the face image needing to verify the identity information into the identity information verification request and sending the identity information verification request to the service processing subsystem;

the service processing subsystem is used for receiving the identity information verification request, extracting the face image contained in the identity information verification request, and acquiring one or more distributed operation servers in the comparison subsystem meeting the load balancing strategy according to the extracted face image and the preset load balancing strategy; splitting the extracted multiple face images according to the acquired distributed operation server, and correspondingly distributing the face images to the acquired distributed operation server; receiving the obtained comparison result information reported by the distributed operation server, and outputting the comparison result information to the query display subsystem for display;

and the comparison subsystem is used for calculating similarity values of the extracted and received features in the face images and the features in a pre-stored feature library to obtain comparison result information, wherein the comparison result information comprises the similarity values and identity information mapped by the similarity values, and each face image corresponds to one piece of comparison result information.

With reference to the second aspect, in a first implementation manner of the second aspect, the query display subsystem includes: a face image acquisition module and a comparison result display module, wherein,

the face image acquisition module is used for acquiring and acquiring a face image and outputting the acquired face image to the service processing subsystem;

and the comparison result display module is used for displaying the comparison result information received from the business processing subsystem.

With reference to the second aspect, in a second implementation manner of the second aspect, the service processing subsystem includes: a central server and a data management server, wherein,

the central server is used for transmitting the face image output by the face image acquisition module to the comparison subsystem, receiving comparison result information returned by the comparison subsystem and respectively outputting the comparison result information to the data management server and the query display subsystem;

and the data management server is used for storing the comparison result information output by the central server.

With reference to the second implementation manner of the second aspect, in a third implementation manner of the second aspect, the service processing subsystem further includes: a dispatch server and a preprocessing server, wherein,

the scheduling server is used for performing distributed scheduling on the face image output by the face image acquisition module according to a preset scheduling strategy and distributing the face image subjected to distributed scheduling to one or more corresponding preprocessing servers; receiving the face image returned from the front processing server and transmitting the face image to the central server;

and the preprocessing server is used for evaluating the quality of the face image distributed by the scheduling server, returning the face image subjected to quality evaluation to the scheduling server, correcting the face image which does not pass quality evaluation according to a preset face image correction algorithm, and returning the corrected face image to the scheduling server.

With reference to the second implementation manner of the second aspect, in a fourth implementation manner of the second aspect, the service processing subsystem further includes:

and the modeling server is used for receiving a modeling task which is output by the scheduling server and contains a modeling face image, extracting the characteristics of the modeling face image, constructing a mapping relation library of the extracted characteristics and the identity information according to the identity information corresponding to the modeling face image, and respectively outputting the mapping relation library to the data management server and the comparison subsystem through the scheduling server and the central server.

With reference to the second aspect, in a fifth implementation manner of the second aspect, the service processing subsystem includes: a load value sorting unit and a loading unit, wherein,

the load value sorting unit is used for acquiring load values of all distributed operation servers with communication connection and sorting the load values in the order from small load value to large load value;

and the loading unit is used for respectively loading the extracted face images to the distributed operation servers according to the sequence of the load values of the distributed operation servers from small to large.

With reference to the fifth implementation manner of the second aspect, in a sixth implementation manner of the second aspect, the loading unit includes: a first load subunit and a second load subunit, wherein,

the first loading subunit is used for selecting the distributed operation server with the minimum load value to load a part of the face images in the extracted face images, so that the load value of the distributed operation server with the minimum load value reaches a preset load threshold value;

and the second loading subunit loads the remaining face images to the distributed operation servers in sequence according to the sequence from small to large of the load values of the distributed operation servers for the distributed operation servers except the distributed operation server with the smallest load value, so that all or part of the load values of the remaining distributed operation servers reach a preset load threshold value until all the face images are loaded.

With reference to the second aspect, in a seventh implementation manner of the second aspect, the modeling server includes: a feature extraction control unit and a feature extraction unit, wherein,

the characteristic extraction control unit is used for receiving a modeling task which is output by the scheduling server and contains a modeling face image, splitting the modeling task according to a load balancing strategy and then distributing the split modeling task to one or more characteristic extraction units which accord with the load balancing strategy; receiving a mapping relation library of the characteristics and the identity information returned by the characteristic extraction unit, and respectively outputting the mapping relation library to a data management server and a comparison subsystem through a scheduling server and a central server after aggregation;

and the feature extraction unit is used for extracting features of the received modeling face image, constructing a mapping relation library of the extracted features and the identity information according to the identity information corresponding to the modeling face image contained in the modeling task, and outputting the mapping relation library to the feature extraction control unit.

With reference to the second aspect, the first implementation manner or the second implementation manner of the second aspect, in an eighth implementation manner of the second aspect, the ratio subsystem includes: a comparison controller and a comparison server, wherein,

the comparison controller is used for receiving the mapping relation database output by the central server, splitting data according to different mapping relations in the mapping relation database to form a plurality of sub-mapping relation databases, and sequentially outputting the plurality of sub-mapping relation databases to corresponding comparison servers, wherein the sub-mapping relation databases stored by different comparison servers are different; receiving the face images output by the central server, and respectively outputting the face images to the comparison server; receiving a preliminary comparison result output by the comparison server, screening to obtain comparison result information, and outputting the comparison result information to the central server;

the comparison server is used for receiving and storing the sub-mapping relation database output by the comparison controller; receiving a face image output by a central server, extracting features, matching the extracted features with a stored sub-mapping relation database, acquiring the top n identity information with the highest similarity with the sub-mapping relation database, and outputting the identity information and the corresponding similarity value as a primary comparison result to a comparison controller.

With reference to the eighth implementation manner of the second aspect, in a ninth implementation manner of the second aspect, the service processing subsystem further includes:

the system configuration server is used for comparing the configuration of the mapping relation between the server and the comparison unit, so that the comparison controller and the comparison server perform corresponding configuration on the server and the comparison unit according to the configuration of the system configuration server.

According to the identity information verification method and system provided by the embodiment of the invention, an identity information verification request is received, wherein the identity information verification request comprises a face image; extracting the face image contained in the identity information verification request, and acquiring one or more distributed operation servers meeting the load balancing strategy according to the extracted face image and a preset load balancing strategy; splitting the extracted multiple face images according to the acquired distributed operation server, and correspondingly distributing the face images to the acquired distributed operation server; receiving comparison result information reported by the obtained distributed operation server, wherein the comparison result information is obtained by the distributed operation server through similarity value calculation of features in extracted and received face images and features in a pre-stored feature library, the comparison result information comprises similarity values and identity information mapped by the similarity values, and each face image corresponds to one piece of comparison result information, so that the identity information verification efficiency can be effectively improved, and the problems that in the existing identity information verification method, the response time of feature matching caused by the fact that the extracted features need to be matched with the pre-set feature library one by one is long, and the identity information verification efficiency is low are solved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a flow chart of an identity information verification method according to an embodiment of the present invention;

fig. 2 is a schematic structural diagram of a second identity information verification system according to an embodiment of the present invention.

Detailed Description

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

It should be understood that the described embodiments are only some embodiments of the invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Fig. 1 is a schematic flow chart of an identity information verification method according to an embodiment of the present invention, and as shown in fig. 1, the method according to the embodiment may include:

step 101, receiving an identity information verification request, wherein the identity information verification request comprises a face image;

in this embodiment, as an optional embodiment, the identity information verification request submitted at one time may include a plurality of face images, for example, 1000 or ten thousand.

In this embodiment, the administrator may actively submit the identity information verification request, or may automatically submit the identity information verification request according to a preset policy, for example, after the camera performs shooting, the identity information verification request is automatically initiated.

In this embodiment, as an optional embodiment, before the receiving the identity information verification request, the method further includes:

collecting an image;

and identifying one or more face images contained in the acquired images according to a preset face identification algorithm, and packaging the identified face images in the identity information verification request.

102, extracting a face image contained in the identity information verification request, and acquiring one or more distributed operation servers meeting a load balancing strategy according to the extracted face image and the preset load balancing strategy;

in this embodiment, as an optional embodiment, the obtaining one or more distributed operation servers meeting the load balancing policy according to the extracted face image and a preset load balancing policy includes:

acquiring load values of all distributed operation servers with communication connection, and arranging the load values in a sequence from small load values to large load values;

In this embodiment, as an optional embodiment, the loading the extracted face images to the distributed arithmetic servers respectively according to the order from small to large of the load value of the distributed arithmetic servers includes:

In this embodiment, the load threshold may be set according to actual needs, for example, may be set as a load when the distributed computing server operates optimally.

In this embodiment, the load balancing policy is to make the load of each distributed operation server not exceed a preset load threshold.

In this embodiment, as another optional embodiment, the obtaining, according to the extracted face image and a preset load balancing policy, one or more distributed operation servers that satisfy the load balancing policy includes:

and sequentially loading the rest face images of the extracted face images from the distributed operation server with the load value smaller than the average load value according to the sequence of the load values of the distributed operation servers from small to large so as to enable the load of the loaded distributed operation server to reach the average load value.

In this embodiment, the load balancing policy is to make the load of each distributed computing server the same or similar.

In this embodiment, as another optional embodiment, after extracting the face image included in the identity information verification request, the face image feature extraction may also be performed on the extracted face image, and according to the extracted face image feature and a preset load balancing policy, one or more distributed operation servers meeting the load balancing policy are obtained.

Step 103, splitting the extracted face image according to the acquired distributed operation server, and correspondingly distributing the face image to the acquired distributed operation server;

in this embodiment, as an optional embodiment, splitting the extracted face image according to the obtained distributed operation server includes:

and extracting the face images loaded by the distributed operation server from the plurality of extracted face images to serve as the splitting packages distributed to the distributed operation server.

In this embodiment, the face images loaded in the distributed operation server are extracted according to the amount of the face images loaded when the distributed operation server is selected, and are distributed to the distributed operation server for corresponding processing.

In this embodiment, as another optional embodiment, if one or more distributed operation servers meeting the load balancing policy are obtained according to the extracted facial image features and a preset load balancing policy, the extracted facial image features are split according to the obtained distributed operation servers and are correspondingly distributed to the obtained distributed operation servers.

In this embodiment, as an optional embodiment, before the splitting the extracted multiple face images according to the obtained distributed computing server, the method further includes:

And 104, receiving comparison result information reported by the acquired distributed operation server, wherein the comparison result information is obtained by the distributed operation server through similarity value calculation of the extracted features in the received face images and the features in a pre-stored feature library, the comparison result information comprises similarity values and identity information mapped by the similarity values, and each face image corresponds to one piece of comparison result information.

In this embodiment, as an optional embodiment, the information of the comparison result is obtained by calculating similarity values between features in the extracted and received face image and features in a feature library stored in advance, and includes:

sorting according to the sequence of similarity values from high to low;

In this embodiment, n is a natural number, and can be set according to actual needs. As an optional embodiment, each face image corresponds to a feature, the feature corresponding to each face image corresponds to a similarity value of n top-ranked bits, and each comparison result information includes n similarity values and identity information to which the n similarity values are respectively mapped.

In this embodiment, as an optional embodiment, the method further includes:

and sequencing the comparison result information of the same person according to the similarity values to obtain identity information mapped by the N similarity values before sequencing.

In this embodiment, N is a natural number, and may be set according to actual needs, and N may be the same as N or different from N.

In this embodiment, as an optional embodiment, the same person may have one or more face images, and the face images and the person have a mapping relationship. Therefore, a plurality of face images of the same person may be distributed to a plurality of distributed operation servers for processing, corresponding to a plurality of comparison result information.

In this embodiment, the identity information verification can be more accurate by aggregating the comparison result information.

In this embodiment, as an optional embodiment, the constructing the pre-stored feature library includes:

In this embodiment, as another optional embodiment, the method further includes:

scanning a bar code on the ticket service to acquire passenger information mapped by the bar code;

and if the passenger information is matched with any identity information in the comparison result information corresponding to the face image of the passenger, confirming that the passenger tickets are consistent.

In the embodiment, the method can be used for verifying the identity of the passenger in accordance with the ticket only with a small amount of labor cost.

In this embodiment, as an optional embodiment, if any identity information in the comparison result information corresponding to the passenger information and the face image of the passenger is not matched, it is determined that the passenger tickets are not consistent, and warning information is issued.

In this embodiment, as an optional embodiment, the method further includes:

and storing the comparison result information.

In the identity information verification method of the embodiment, an identity information verification request is received, wherein the identity information verification request comprises a face image; extracting the face image contained in the identity information verification request, and acquiring one or more distributed operation servers meeting the load balancing strategy according to the extracted face image and a preset load balancing strategy; splitting the extracted multiple face images according to the acquired distributed operation server, and correspondingly distributing the face images to the acquired distributed operation server; and receiving comparison result information reported by the acquired distributed operation server, wherein the comparison result information is obtained by the distributed operation server through similarity value calculation of the extracted features in the received face images and the features in a pre-stored feature library, the comparison result information comprises similarity values and identity information mapped by the similarity values, and each face image corresponds to one piece of comparison result information. Therefore, through distributed deployment, the face image recognition is processed in a distributed mode, the response time of identity information verification is effectively shortened, the identity information verification efficiency is improved, and the functions of rapid face image comparison and identity information recognition with large library capacity can be realized.

Fig. 2 is a schematic structural diagram of a second identity information verification system according to an embodiment of the present invention. As shown in fig. 2, the identity information verification system of the present embodiment may include: a query display subsystem, a service processing subsystem and a comparison subsystem, wherein,

in this embodiment, the query display subsystem is connected to the service processing subsystem via the internet.

In this embodiment, as an optional embodiment, the query display subsystem includes: a face image acquisition module and a comparison result display module, wherein,

in this embodiment, the face image acquisition module may acquire and obtain the face image according to the received face image acquisition instruction, or may automatically acquire and obtain the face image.

In this embodiment, as an optional embodiment, the acquiring and obtaining the face image by the face image acquisition module includes:

collecting an image;

and identifying one or more face images contained in the acquired images according to a preset face identification algorithm.

In this embodiment, after the image is acquired, the image may be preprocessed, for example, the image with higher quality per second may be selected by an algorithm to perform face recognition, which may improve the comparison quality and save resources. For example, one face image with high quality is selected from 25 face images acquired by the acquisition module per second, that is, the image with high quality per second is subjected to face recognition according to a preset face recognition algorithm.

In this embodiment, as an optional embodiment, the comparison result information is specific identity information corresponding to the passenger (the face image) and a corresponding similarity value. As another optional embodiment, the comparison result information is specific identity information corresponding to the passenger (face image), for example, result information indicating whether the passenger identity (face image) is consistent with the ticket; alternatively, the passenger identity (face image) is a suspicious evasion; alternatively, the passenger identity (facial image) is the missing population; or the passenger identity (face image) is a frequent visitor, etc.

In this embodiment, as an optional embodiment, the query display subsystem further includes:

and the face image warehousing management module is used for storing the comparison result information received from the business processing subsystem.

In this embodiment, as an optional embodiment, the service processing subsystem includes: a central server and a data management server, wherein,

In this embodiment, the data management server may facilitate subsequent history query and output a history query record by storing the comparison result information.

In this embodiment, as another optional embodiment, the service processing subsystem further includes: a dispatch server and a preprocessing server, wherein,

In this embodiment, the preprocessing server may provide a high-quality face image, so as to facilitate subsequent feature comparison.

In this embodiment, as a further optional embodiment, the service processing subsystem further includes:

In this embodiment, as an optional embodiment, there may be one or more modeling servers, and if there are multiple modeling servers, the scheduling server is further configured to distribute, through a load balancing policy, the modeling tasks to the one or more modeling servers before outputting the modeling tasks, and output corresponding modeling subtasks to the assigned modeling servers. Certainly, in practical application, the scheduling server may also output the modeled face image to the preprocessing server for quality evaluation, and then the scheduling server distributes the modeled face image returned by the preprocessing server according to a load balancing policy.

In this embodiment, the modeled face image is a face image with known identity information. One modeling face image can correspond to one or more identity information, for example, the modeling face image can correspond to any one or more combinations of passenger identity information, escapement identity information, missing population identity information, visiting person identity information and labor and education person identity information. As another alternative, an identity information of a person may also correspond to multiple modeled face images.

In this embodiment, as an optional embodiment, in the constructed mapping relation library, one feature maps one or more identity information, and one identity information maps one or more features.

In this embodiment, as an optional embodiment, the modeling task refers to a task of extracting features from a face image, for example, the extraction of features from a face image indicates that a modeling task is completed.

In this embodiment, as an optional embodiment, the load balancing policy may be a round-robin distribution processing policy, or may be a policy that determines whether the modeling server is in a busy state according to information such as a CPU, a memory, and a task queue length of the modeling server, and parameters such as a response time of the modeling server, and preferentially distributes the modeling task to the modeling server that is not busy.

In this embodiment, as another optional embodiment, the modeling server may also perform face image feature extraction on each of the plurality of face images in the identity information verification request, and output the extracted face image features to the scheduling server, so that the scheduling server performs scheduling distribution on the face image features.

In this embodiment, as an optional embodiment, the modeling server includes: a feature extraction control unit and a feature extraction unit, wherein,

the characteristic extraction control unit is used for receiving a modeling task which is output by the scheduling server and contains a modeling face image, splitting the modeling task according to a load balancing strategy and then distributing the split modeling task to one or more characteristic extraction units which accord with the load balancing strategy; receiving a mapping relation library of the characteristics and the identity information returned by the characteristic extraction unit, and respectively outputting the mapping relation library to a data management server and a comparison subsystem through a scheduling server and a central server after aggregation, wherein the aggregation refers to integrating data according to preset rules and business logic;

In this embodiment, when the number of the modeled face images put in storage is large, for example, three or more than ten thousand, since the number of the modeled face images supported by one process is limited, and a plurality of processes are required for processing, the feature extraction control unit selects the feature extraction unit according to the load balancing policy, and the feature extraction and the mapping relationship construction can be completed quickly by adopting the distributed deployment of the plurality of feature extraction units.

In this embodiment, as an optional embodiment, the mapping relationship library includes: passenger storehouses, prisoner storehouses, population missing storehouses, visiting personnel storehouses, labor and education personnel storehouses and the like. Wherein, one characteristic of people can be respectively mapped to a passenger library, an evasion library, a population missing library and the like, and the identity information of one person in the passenger library can also be mapped to a plurality of characteristics of the person. The mapping relation library can be stored as external data, and the external data is managed by using the data management server.

In this embodiment, as an optional embodiment, the service processing subsystem includes: a load value sorting unit and a loading unit, wherein,

In this embodiment, as an optional embodiment, the loading unit includes: a first load subunit and a second load subunit, wherein,

and the second loading subunit is used for sequentially selecting the distributed operation servers with other load values to load the rest face images according to the loading mode of the distributed operation server with the minimum load value and the sequence of the load values of the distributed operation servers from small to large. .

In this embodiment, as an optional embodiment, the system further includes:

and the system configuration server is used for comparing the configuration of the mapping relation of the comparison server in the subsystem, so that the comparison controller in the comparison subsystem carries out corresponding configuration on the server by comparison according to the configuration of the system configuration server.

In this embodiment, the system configuration server may configure the comparison controller to split or not split the mapping relationship, so that the comparison server stores the mapping relationship corresponding to the split or not split, and for the case that the comparison server stores the mapping relationship not split, the face image is distributed according to the load balancing policy, and for the case that the comparison server stores the split mapping relationship, the face image is distributed to each comparison server.

In this embodiment, as an optional embodiment, the system configuration server, the central server, and the data management server form a data management system.

In this embodiment, as an optional embodiment, the subsystem includes: a comparison controller and a comparison server, wherein,

the comparison controller is used for receiving the mapping relation database output by the central server, splitting data by different mapping relations in the mapping relation database to form a plurality of sub-mapping relation databases, and sequentially outputting the plurality of sub-mapping relation databases to corresponding comparison servers, wherein the sub-mapping relation databases stored by different comparison servers are different; receiving the face images output by the central server, and respectively outputting the face images to the comparison server; receiving a preliminary comparison result output by the comparison server, screening to obtain comparison result information, and outputting the comparison result information to the central server;

In this embodiment, as an optional embodiment, the comparison server may not store the sub-mapping relationship database in advance, and when the extracted features are matched, the corresponding sub-mapping relationship database is loaded from the data management server to perform matching.

In this embodiment, the sub-mapping relationship database includes a mapping relationship between features and identity information, the similarity values of the extracted features are obtained according to the sub-mapping relationship database, the obtained similarity values are sorted from large to small, the first n-bit similarity values are selected, the identity information of the feature mapping in the sub-mapping relationship database respectively corresponding to the first n-bit similarity values is obtained, and the identity information and the corresponding similarity values are obtained. For example, if the sub-mapping relationship database stored in the comparison server is a passenger library, the passenger library includes 1000 features, and 1000 pieces of passenger identity information are respectively mapped, and 10 received face images are received, similarity values between the extracted features of each face image and the 1000 features are calculated, 5 pieces of similarity values with the top rank are extracted, passenger identity information of feature mapping of the 5 similarity values in the sub-mapping relationship database is obtained, that is, the features of each face image correspond to the 5 similarity values and corresponding passenger identity information, and a preliminary comparison result is formed, and the features of 10 face images correspond to the 50 similarity values and corresponding passenger identity information.

In this embodiment, as an optional embodiment, the comparison controller may also not split the received mapping relationship library, and directly output the mapping relationship libraries to each comparison server for storage, and after receiving the face image output by the central server, the comparison controller allocates the face image to one or more comparison servers conforming to the load balancing policy according to the load balancing policy to process, so that the one comparison server processes a part of the face image.

In this embodiment, as an optional embodiment, the comparison server includes: a comparison control unit and a comparison unit, wherein,

the number of the comparison control units is one, and the number of the comparison units is one or more.

The comparison control unit is used for splitting the received sub-mapping relation databases and then sequentially outputting the split sub-mapping relation databases to the corresponding comparison units, and the split sub-mapping relation databases stored in different comparison units are not overlapped with each other; respectively outputting the received face images to a comparison unit; receiving an initial comparison result output by the comparison unit, screening to obtain a preliminary comparison result, and outputting the preliminary comparison result to the comparison controller;

and the comparison unit is used for extracting the features of the received face image, matching the extracted features with the stored split sub-mapping relation database, acquiring the first n identity information with the highest similarity with the split sub-mapping relation database, and outputting the identity information and the corresponding similarity value as an initial comparison result to the comparison server.

In this embodiment, as an optional embodiment, the comparison control unit may also not split the received sub-mapping relationship database or the mapping relationship database, directly output the mapping relationship database or the sub-mapping relationship database to each comparison unit for storage, and after receiving the face image, allocate the face image to one or more comparison units conforming to the load balancing policy according to the load balancing policy for processing.

In this embodiment, each comparison unit may quickly complete the comparison task in blocks, after the comparison is completed, the comparison result is returned to the comparison server, the comparison server summarizes and screens the comparison result, and then sends the comparison result to the comparison controller, after the comparison controller receives the comparison result, the comparison controller screens again and then sends the comparison result to the central server, and the central server transmits the comparison result to the database (data management server) and the scheduling server, and sends the comparison result to the query display subsystem through the scheduling server to be displayed.

In this embodiment, large-scale data (face images) are segmented and dispersed into a plurality of comparison units for processing.

In this embodiment, as an optional embodiment, the comparison server may manage one or more comparison units, where the comparison server and the comparison units are both processes, and the comparison unit notifies the comparison server through the configuration item and notifies the number of the comparison units, where each comparison unit has a unique identifier (Id), the comparison unit performs a specific comparison operation, and the comparison server is responsible for scheduling and managing the comparison units.

In this embodiment, the system configuration server is specifically configured to compare the mapping relationship between the server and the comparison unit, so that the comparison controller and the comparison server perform corresponding configuration on the server and the comparison unit according to the configuration of the system configuration server. The system configuration server can configure the comparison controller to split or not split the mapping relationship, so that the comparison server stores the mapping relationship corresponding to the split or not split, and for the situation that the comparison server stores the mapping relationship not split, the face image is distributed according to the load balancing strategy, and for the situation that the comparison server stores the split mapping relationship, the face image is distributed to each comparison server.

In this embodiment, the system may further include a public Interface subsystem, where the public Interface subsystem includes a public Interface server, an input Interface of the public Interface subsystem is connected to an Application Programming Interface (API) of another system, and an output Interface of the public Interface subsystem is connected to the central server, so that the API that can be quickly accessed by a third-party system is provided, and information sharing with the third-party system is implemented. For example, any platform or system (other video surveillance platform or other authentication system) that implements face recognition functionality may be accessed.

The system of this embodiment may be used to implement the technical solution of the method embodiment shown in fig. 1, and the implementation principle and the technical effect are similar, which are not described herein again.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. The term "comprising", without further limitation, means that the element so defined is not excluded from the group consisting of additional identical elements in the process, method, article, or apparatus that comprises the element.

All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments.

In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.

The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.

It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof.

In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.

It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.

For convenience of description, the above systems are described separately with the functions divided into various units/modules. Of course, the functionality of the units/modules may be implemented in one or more software and/or hardware implementations of the invention.

From the above description of the embodiments, it is clear to those skilled in the art that the present invention can be implemented by software plus necessary general hardware platform. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which may be stored in a storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method according to the embodiments or some parts of the embodiments.

The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. An identity information verification method, comprising:

receiving an identity information verification request, wherein the identity information verification request comprises a face image; the same person has a plurality of face images, and the face images and the person have a mapping relation;

splitting the extracted multiple face images according to the acquired distributed operation server, and correspondingly distributing the face images to the acquired distributed operation server; distributing a plurality of face images of the same person to a plurality of distributed operation servers for processing, wherein the face images correspond to a plurality of comparison result information;

receiving comparison result information reported by the acquired distributed operation server, and performing aggregation processing on the result information by comparison, wherein the comparison result information is obtained by the distributed operation server through similarity value calculation of extracted features in the received face image and features in a pre-stored feature library, the comparison result information comprises a similarity value and identity information mapped by the similarity value, and each face image corresponds to one piece of comparison result information;

the splitting of the extracted multiple face images according to the acquired distributed operation server comprises the following steps: and extracting the face images loaded by the distributed operation server from the plurality of extracted face images to serve as the splitting packages distributed to the distributed operation server.

2. The identity information verification method of claim 1, further comprising:

collecting images, identifying one or more face images contained in the collected images according to a preset face identification algorithm, and packaging the identified face images in the identity information verification request; and the number of the first and second groups,

and displaying the received comparison result information.

3. The identity information verification method of claim 1, further comprising:

and storing the comparison result information.

4. The identity information verification method according to claim 3, wherein before the splitting of the extracted plurality of face images according to the acquired distributed computing server, the method further comprises:

5. The identity information verification method of claim 3, wherein constructing the pre-stored feature library comprises:

6. The identity information verification method according to claim 1, wherein the obtaining one or more distributed operation servers that satisfy the load balancing policy according to the extracted face image and a preset load balancing policy comprises:

7. The identity information verification method according to claim 6, wherein the loading the extracted face images to the distributed computation servers in the order of the load values of the distributed computation servers from small to large respectively comprises:

8. The identity information verification method according to claim 1, wherein the obtaining one or more distributed operation servers that satisfy the load balancing policy according to the extracted face image and a preset load balancing policy comprises:

and sequentially loading the rest face images of the extracted face images from the distributed operation server with the load value smaller than the average load value according to the sequence of the load value of the distributed operation server from small to large so as to enable the load value of the loaded distributed operation server to reach the average load.

9. The identity information verification method according to any one of claims 1 to 3, wherein the comparison result information is obtained by calculating similarity values of the extracted features in the received face image and the features in a pre-stored feature library, and includes:

sorting according to the sequence of similarity values from high to low;

10. An identity information verification system, comprising: a query display subsystem, a service processing subsystem and a comparison subsystem, wherein,

the query display subsystem is used for packaging the face image needing to verify the identity information into the identity information verification request and sending the identity information verification request to the service processing subsystem; the same person has a plurality of face images, and the face images and the person have a mapping relation;

the service processing subsystem is used for receiving the identity information verification request, extracting the face image contained in the identity information verification request, and acquiring one or more distributed operation servers in the comparison subsystem meeting the load balancing strategy according to the extracted face image and the preset load balancing strategy; splitting the extracted multiple face images according to the obtained distributed operation server, and correspondingly distributing the split multiple face images to the obtained distributed operation server, wherein the multiple face images of the same person are distributed to the multiple distributed operation servers for processing and correspond to multiple comparison result information; receiving the obtained comparison result information reported by the distributed operation server, comparing the comparison result information, performing aggregation processing on the result information, and outputting the result information to the query display subsystem for display;

the comparison subsystem is used for calculating similarity values of the extracted and received features in the face images and the features in a pre-stored feature library to obtain comparison result information, wherein the comparison result information comprises the similarity values and identity information mapped by the similarity values, and each face image corresponds to one piece of comparison result information;

wherein the service processing subsystem is specifically configured to: and extracting the face images loaded by the distributed operation server from the plurality of extracted face images to serve as the splitting packages distributed to the distributed operation server.

11. The identity information verification system of claim 10, wherein the query display subsystem comprises: a face image acquisition module and a comparison result display module, wherein,

12. The identity information verification system of claim 10, wherein the service processing subsystem comprises: a central server and a data management server, wherein,

13. The identity information verification system of claim 12, wherein the service processing subsystem further comprises: a dispatch server and a preprocessing server, wherein,

14. The identity information verification system of claim 12, wherein the service processing subsystem further comprises:

15. The identity information verification system of claim 10, wherein the service processing subsystem comprises: a load value sorting unit and a loading unit, wherein,

16. The identity information verification system of claim 15, wherein the loading unit comprises: a first load subunit and a second load subunit, wherein,

and the second loading subunit is used for sequentially selecting the distributed operation servers with other load values to load the rest face images according to the loading mode of the distributed operation server with the minimum load value and the sequence of the load values of the distributed operation servers from small to large.

17. The identity information verification system of claim 14, wherein the modeling server comprises: a feature extraction control unit and a feature extraction unit, wherein,

18. The identity information verification system of any one of claims 10 to 12, wherein the alignment subsystem comprises: a comparison controller and a comparison server, wherein,

19. The identity information verification system of claim 18, wherein the service processing subsystem further comprises: