Security and protection system based on face recognition

Abstract

The invention relates to the field of security and protection, and discloses a security and protection system based on face recognition, which is mainly composed of a block chain network plurality of terminal equipment face data and identity database information processing module and an access control management platform. The block chain network realizes distributed storage and synchronous verification of face recognition data and access control event records, and each terminal device has data writing and reading functions as a node of the block chain network. The face data and identity database is deployed in part of terminal nodes in a distributed mode, data synchronization is achieved through a block chain network, and the information processing module is responsible for verification and version control of received data. The access control management platform comprises a plurality of function modules, when the system is applied to scenes such as a community, efficient access control management can be achieved through face recognition, data safety and reliable evidence storage are guaranteed by means of the block chain technology, meanwhile, the abnormal event processing efficiency and the user experience are improved, and compared with a traditional security and protection system, the system has obvious advantages.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of security and defense, more particularly to a security and defense system based on face recognition. BACKGROUND

[0002] With the acceleration of urbanization, the population of residential areas is increasing, and the flow of personnel is becoming more frequent. The traditional security and defense system gradually exposes many problems in access control management and security protection. The traditional access control system mostly uses card swiping, password and other methods, which has security risks such as card loss and password leakage, and cannot effectively identify the situation of using someone else's identity. At the same time, the data storage of these systems mostly adopts a centralized architecture. Once the central server fails or is attacked, it is easy to cause data loss or tampering, which brings risks to the security of the residential area. In addition, the management of face data and access event records in the traditional system is relatively scattered, and the data synchronization between different terminal devices is not timely. When there are visitors or resident information updates, it is easy to have inconsistent information, which affects the access efficiency and the effect of security and defense management. Moreover, in terms of abnormal event handling, the traditional system lacks an effective linkage mechanism. After the identification fails, it often needs to be verified one by one manually, which is slow in response and difficult to quickly respond to potential security threats. In addition, the visitor management process is complicated and needs to be registered manually, which not only takes a long time, but also may have information recording errors or omissions, which brings inconvenience to the security management of the residential area. SUMMARY

[0003] In view of the deficiencies in the prior art, the purpose of the present application is to provide a security and defense system based on face recognition to solve the problem that the traditional security and defense inspection process of the city residential area cannot effectively identify the identity of others, and to improve the security level of the security and defense system.

[0004] To achieve the above purpose, the present application provides the following technical scheme:

[0005] A security and defense system based on face recognition, comprising:

[0006] a blockchain network for realizing distributed storage and synchronous verification of face recognition data and access event records, the blockchain network adopting a consortium chain architecture;

[0007] a plurality of terminal devices, including a face recognition access control terminal and a management terminal, each terminal device serving as one of the nodes in the blockchain network and having data writing and reading functions;

[0008] A face data and identity database is distributed in some terminal nodes, and is configured to store face feature information and identity information of residents and visitors in a cell. The database is configured with a first data packet for data synchronization. Each database synchronizes encrypted data content to other nodes through a blockchain network at a preset time interval or in an event triggered manner.

[0009] An information processing module is configured with a second data packet, and is configured to verify the structural integrity and encryption consistency of the received first data packet, and to perform version control and update confirmation through a smart contract.

[0010] A gate management platform is configured to perform unified scheduling and policy execution of identity data recording, face recognition, abnormal response and security linkage.

[0011] Further, the gate management platform includes an identity information management module, a gate monitoring module, an abnormal processing module and a security response module.

[0012] The identity information management module is configured to input and manage the identity information and face data of residents and visitors.

[0013] The gate monitoring module is configured to collect personnel face images and compare them with face features in the database.

[0014] If the identification is successful, an open door signal is output, and if the identification fails, the record is uploaded to the blockchain.

[0015] The abnormal processing module is configured to start a manual identification process when the number of identification failures exceeds a threshold.

[0016] The security response module is configured to generate an abnormal alarm and notify security personnel to handle it.

[0017] Further, the identity information management module includes:

[0018] A data entry unit is configured to receive and format the identity information and face images of residents or visitors in the cell.

[0019] A resident file subunit is configured to store the identity information and face feature vectors of long-term residents, and associate them with the house number.

[0020] A visitor binding subunit is configured to establish a binding relationship between the identity information of temporary visitors and the collected face images, and generate a temporary access credential.

[0021] A visitor file subunit is configured to save the bound visitor face and identity record, and set an effective access time window.

[0022] A data chaining control unit is configured to upload the newly added or changed information in the resident profile and the visitor profile to the blockchain network through a first data packet, so as to realize node synchronization.

[0023] Further, the identity information management module further comprises an information updating unit configured to acquire the face image collected by the access control monitoring module, and compare the face image with historical face data in terms of difference degree, and if the difference degree exceeds a preset threshold, automatically trigger image updating and upload to the blockchain network to replace the old data.

[0024] Further, the abnormality processing module comprises:

[0025] An interruption detection unit is configured to determine whether the same face image is collected multiple times within a preset time period and is not matched successfully.

[0026] An artificial processing module is configured to receive abnormal image and video data, and perform artificial identity verification to determine whether it is a registered user or a misrecognition behavior.

[0027] An abnormality chaining confirmation subunit is configured to pack the artificial processing result, the corresponding image data, the operation record and the identity of the reviewer into an abnormality processing data block, and write the processing result into the blockchain network through a preset abnormality event chaining contract.

[0028] The abnormality processing module is further configured with a threshold dynamic adjustment strategy, which comprises dynamically adjusting the threshold of the number of recognition failures in combination with the frequency of historical abnormal event types on the chain.

[0029] Further, the artificial processing module further comprises an error recognition strategy unit, which is configured to determine the recognition failure reason by image acquisition conditions, including:

[0030] Acquire the illumination parameter, saturation and color temperature value of the current image and the database image, and determine the environmental difference degree by using the image brightness histogram difference and principal component comparison;

[0031] When the difference degree index exceeds a preset threshold, generate an auxiliary identification light adjustment suggestion, including parameters such as light direction, light intensity level, light source color temperature, and synchronize to the corresponding access control terminal control system;

[0032] The auxiliary identification light adjustment suggestion is uploaded to the blockchain as auxiliary identification suggestion data packet.

[0033] Further, the access control management platform further comprises an abnormality recognition and security linkage module, which comprises:

[0034] An identity association monitoring unit is configured to confirm whether a face identity is successfully matched in each face recognition comparison process, and add an unbound identity code in a recognition failure record if the face identity cannot be matched;

[0035] A behavior pattern recognition subunit is configured to construct an abnormal behavior pattern vector based on continuous recognition failures of the same terminal, attempts to enter or exit at night, multi-target recognition failures of the same person, and the like, and compare the abnormal behavior pattern vector with an existing behavior model on the chain;

[0036] A security warning pushing unit is configured to submit an event as an abnormal event data block to a blockchain network and synchronously push the event to a security personnel terminal when an abnormal pattern is triggered, and a warning content includes a terminal position, an image snapshot, a behavior description, and a suggested response operation;

[0037] The module supports linkage calling with an abnormal behavior event library on the chain to form community-level abnormal type trend statistics and automatic early warning rule updating.

[0038] Further, the access control management platform further includes a visitor permission management module, and the visitor permission management module includes:

[0039] A visitor registration subunit is configured to receive identity information, a face image, and an access reservation time of a visitor through a two-dimensional code, a self-service terminal, or a property APP;

[0040] A visitor input subunit is configured to input the above information into a visitor feature library after standardization processing, and set an effective access time and a target resident;

[0041] An authorization generation subunit is configured to generate a corresponding visitor access credential, and synchronously submit the visitor access credential to a blockchain network to ensure a trusted record of an authorization chain, and generate an authorization record on-chain credential hash identification for subsequent verification;

[0042] The visitor permission management module updates a visitor access state on the chain.

[0043] Further, the dynamic information linkage and display module includes:

[0044] An information recognition association unit is configured to read a chain identity hash value corresponding to an identification event after successful access control identification, and call corresponding visitor or resident archive data on the chain;

[0045] An information pushing processing unit is configured to determine whether there is un-read visitor record, express delivery visit, access record change, and the like based on the recognized identity, and generate corresponding notification data according to types;

[0046] An interactive display unit is configured to display the pushing information in a form of text, image, or voice on an access control screen, a property front desk, or a mobile application end;

[0047] The push synchronization subunit is configured to submit each push record to the blockchain network in an unalterable format.

[0048] Further, the blockchain network is constructed based on a consortium chain, each terminal node has data writing right, and the following mechanisms are adopted by the blockchain network to guarantee data consistency and credible evidence storage:

[0049] Node writing rule: each writing needs to contain a node signature, an event type code, a structured data digest and a timestamp;

[0050] Smart contract rule library: including a successful contract identification contract, a failed contract identification contract, an exception handling contract, a visitor authorization contract and a data updating contract;

[0051] Synchronization control logic: a synchronization strategy based on polling and event triggering is adopted, and the on-chain smart contract is compared according to the version number comparison rule and the preset synchronization period;

[0052] Historical audit mechanism: all face recognition events, access records, abnormal responses and the like form multi-dimensional identification fields on the chain.

[0053] Compared with the prior art, the security system based on face recognition has the following beneficial effects:

[0054] Compared with the traditional security system, the security system based on face recognition has the following beneficial effects. From the perspective of data security, the system adopts the consortium chain architecture of the blockchain network to realize distributed storage and synchronous verification of data, avoiding the risk of data loss or tampering in centralized storage. All face recognition data, access event records and manual processing results are stored and evidenced through the blockchain, and each node participates in data verification, ensuring the authenticity and non-tamperability of data, greatly improving the security and credibility of data.

[0055] In terms of management efficiency, the system realizes the distributed deployment of face data and identity database and the data synchronization mechanism, ensuring the consistency of data of each terminal node. When a new resident moves in or a visitor registers, the relevant information can be quickly synchronized to each node through the blockchain network, avoiding recognition errors caused by information lag. At the same time, the access management platform realizes the unified scheduling of identity data records, face recognition, abnormal response and security linkage, reducing the manual intervention link. For example, after a visitor makes an appointment through a property APP, the system can automatically complete the information input, authorization generation and other processes without manual registration, improving the visitor management efficiency.

[0056] In terms of abnormality processing and security linkage, the system has perfect abnormality recognition and response mechanism. The abnormality processing module can start the manual recognition process in time when the number of recognition failures exceeds the threshold value, and the threshold value can be dynamically adjusted according to the historical abnormality event frequency, thereby improving the timeliness of abnormality event processing. The security response module can quickly push the warning information to the security personnel when the abnormality mode is triggered, thereby realizing the rapid linkage of security and defense. For example, when a stranger tries to enter the community for many times, the system can quickly inform the security personnel to handle it on the spot, thereby reducing the security risk.

[0057] In terms of user experience, residents do not need to carry access control cards and can quickly pass through the face recognition, thereby avoiding the trouble of forgetting to carry the card. The dynamic information linkage and display module can push the information such as express delivery visit after the resident recognition is successful, thereby improving the life convenience of the residents. At the same time, the system analyzes the recognition failure reason through the error recognition strategy unit and adjusts the access control terminal light parameters, thereby reducing the recognition failure caused by environmental factors, improving the success rate of face recognition, and making the residents and visitors pass through more smoothly.

[0058] In addition, the historical audit mechanism of the blockchain provides traceable basis for community security management. The management personnel can understand the past access control events and abnormality processing conditions through the query of on-chain records, thereby facilitating the summary and optimization of security management. The collaborative work of each terminal node under the alliance chain architecture also reduces the operation and maintenance cost of the system, thereby improving the stability and reliability of the system. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 FIG. 1 is a schematic diagram of the overall structure of the system in the embodiment;

[0060] Figure 2 FIG. 2 is a function block diagram of the access control management platform module in the embodiment;

[0061] Figure 3 FIG. 3 is a data flow diagram of face recognition and blockchain in the embodiment;

[0062] Figure 4 FIG. 4 is a flowchart of abnormality processing and security linkage in the embodiment;

[0063] Figure 5 FIG. 5 is a schematic diagram of visitor permission management and dynamic information display in the embodiment. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings of the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0065] It should be understood that when an element, referred to as being "on" another element, it can be directly on the other element or intervening elements can also be present. In addition, it should be understood that when an element is referred to as being "connected", "coupled", or "disposed" to another element, it can be directly connected, coupled, or disposed to the other element or intervening elements can also be present. In addition, it should be understood that when an element is referred to as being "disposed" to another element, it can be directly disposed on the other element or intervening elements can also be present. The terms "vertical", "horizontal", "left", "right", and similar expressions as used herein are for illustrative purposes only.

[0066] 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 application belongs. The terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0067] Please refer to Figures 1 to 5 The embodiment provides a security system based on face recognition. The security system based on face recognition is a security solution specially designed for community and other scenes. The security system combines face recognition technology and blockchain technology to realize efficient and safe access control management and security linkage. The components and working mechanism of the system are described in detail below.

[0068] First, the blockchain network plays a role in the entire system to realize distributed storage and synchronous verification of face recognition data and access control event records. The blockchain network here adopts a consortium chain architecture. The consortium chain architecture refers to a blockchain that is jointly managed by multiple pre-authorized institutions or nodes. There is a certain trust basis between these nodes, such as community property management, security companies, and other related parties as authorized nodes participating in the system. Only authorized nodes can join the network and write and verify data, which ensures data security and improves management efficiency.

[0069] A number of terminal devices include face recognition access control terminals and management terminals. Each terminal device is a node in the blockchain network and has data writing and reading functions. The face recognition access control terminal can be a device installed at the entrance of a community for collecting and identifying face images. The management terminal can be a computer or other device used by property management personnel for system management and operation. These terminals as blockchain nodes can write relevant data to the blockchain when access control events occur or face data is updated, and can also read the required data from the blockchain, such as management personnel querying access control records in the community through the management terminal.

[0070] The face data and identity database is distributed in some terminal nodes, such as some face recognition access control terminals and management terminals, which store the database. This database is used to store the face feature information and identity information of residents and visitors in the community. The face feature information of residents can be a face feature vector extracted by a specific algorithm, and the identity information includes name, ID number, etc. The information of visitors is similar, and also contains access-related information. The first data packet configured in the database is a data set for data synchronization, which contains face data and identity information that need to be synchronized. Each database will synchronize the encrypted data content to other nodes through the blockchain network at a preset time interval, such as every hour, or in an event-triggered manner, such as when new residents move in and enter data. Encryption processing can use common encryption algorithms, such as the AES encryption algorithm, to ensure that data is not leaked during transmission.

[0071] The information processing module is configured with a second data packet, which is a data set for verifying the first data packet and contains relevant information and rules required for verification. When the information processing module receives the first data packet, it will verify the structural integrity, check whether the data format is correct and the content is complete, such as whether the resident's ID number information is missing, and at the same time, perform encryption consistency verification to confirm that the received data encryption is consistent with that at the time of sending, to avoid data tampering during transmission. Then, through the smart contract, version control and update confirmation are performed. The smart contract is a pre-written computer program that judges whether the updated version is legal when the data needs to be updated, and allows the update after confirmation and records the relevant information.

[0072] The access control management platform is responsible for unified scheduling and policy execution of identity data recording, face recognition, abnormal response, and security linkage. For example, when someone performs face recognition at the access control terminal, the access control management platform will dispatch relevant modules for face comparison, and if the identification is abnormal, it will start the corresponding abnormal response strategy and notify the security personnel.

[0073] The access control management platform comprises an identity information management module, an access control monitoring module, an abnormality processing module and a security response module. The identity information management module is used for inputting and managing the identity information and face data of residents and visitors. For example, when a resident moves in, the property personnel inputs the resident's ID information and collects the face image through the module; when a visitor visits, the relevant information of the visitor is inputted. The access control monitoring module is used for collecting face images and comparing them with the face features in the database. It collects face images through a camera, then extracts features and compares them with the features in the database. If the identification is successful, an opening signal is outputted, and the access control is opened; if the identification fails, the relevant information of the event, such as the failure time and the collected face image, is recorded and uploaded to the blockchain for storage. The abnormality processing module is used for starting the manual identification process when the number of identification failures exceeds the threshold value. The threshold value can be set to 3 times. When the same person fails to be identified for 3 times in succession, the system notifies the property management personnel to perform manual identification, and the management personnel judges by checking the collected image information. The security response module is used for generating an abnormality warning and notifying the security personnel to handle it. For example, when an abnormal situation of multiple identification failures occurs, a warning information is generated and sent to the mobile phone of the security personnel through the system, informing the location and relevant situation of the abnormality.

[0074] The identity information management module comprises a data input unit, a resident file subunit, a visitor binding subunit, a visitor file subunit and a data chain control unit. The data input unit is used for receiving and formatting the identity information and face image of the residents or visitors in the community, such as storing the received ID number information in a fixed format and standardizing the face image to meet the database storage requirements. The resident file subunit is used for storing the identity information and face feature vector of long-term residents and associating them with the house number. The house number can be a unique number allocated by the community to each household, such as the house number of 301 in unit 2 of building 1 is 12301, so the corresponding resident information can be quickly queried through the house number. The visitor binding subunit is used for establishing a binding relationship between the identity information of temporary visitors and the collected face image, and generating a temporary access credential. The temporary access credential is a unique code used to identify the access rights of the visitor. The visitor file subunit is used for saving the bound visitor face and identity record and setting an effective access time window, such as the visitor access time is from 9 am to 11 am, so the visitor can enter the community through face recognition within this time window. The data chain control unit is used for uploading the newly added or changed information in the resident file and visitor file to the blockchain network through the first data packet, realizing node synchronization, such as after a new resident moves in and the information is inputted, the unit will package the new information into the first data packet and upload it to the blockchain, and other nodes receive it for synchronous storage.

[0075] The identity information management module further comprises an information updating unit configured to obtain the face image collected by the access control monitoring module and compare the difference degree with the historical face data. The difference degree is obtained by calculating the distance between the two face feature vectors, for example, the Euclidean distance can be used for calculation, assuming that the historical face feature vector is X (x1, x2,..., xn) and the newly collected face feature vector is Y (y1, y2,..., yn), and the Euclidean distance formula is where n is the dimension of the feature vector, which is 128 here. Operation example: assuming that the first two elements of the historical feature vector X are 1.2 and 0.8, the first two elements of the new feature vector Y are 1.5 and 1.0, and the other elements are the same and the difference is 0, then The difference degree threshold is set to 0.5, and if the calculated difference degree exceeds the threshold, the image is automatically updated and uploaded to the blockchain network to replace the old data. For example, a resident previously entered face data without glasses, and later wore glasses for a long time, resulting in a difference degree between the collected face image and the historical data exceeding the threshold, so the system automatically updates the resident's face data.

[0076] The abnormality processing module comprises an interruption detection unit, an artificial processing module and an abnormality on-chain confirmation subunit. The interruption detection unit is configured to determine whether the same face image is collected multiple times within a preset time period and is not matched successfully. The preset time period can be set to 5 minutes, and if the same face is collected 3 times within 5 minutes and is not matched successfully, it will be detected by the unit. The artificial processing module is configured to receive abnormal image and video data, perform artificial identity verification, and determine whether it is a registered user or a misidentification behavior, for example, the property management personnel determines whether the person is a community resident but fails to be recognized due to facial changes or an outsider by checking the relevant images and videos. The abnormality on-chain confirmation subunit is configured to package the artificial processing result, the corresponding image data, the operation record and the identity of the auditor into an abnormality processing data block, and write the processing result into the blockchain network through a preset abnormal event on-chain contract, so that the processing result is permanently stored and cannot be tampered with, facilitating subsequent queries. The abnormality processing module is further configured with a threshold dynamic adjustment strategy, which dynamically adjusts the recognition failure times threshold in combination with the frequency of historical abnormal event types on the chain. For example, if the number of attempts to enter the community by outsiders increases in a certain time period, the frequency of recognition failure events in the historical abnormal events increases, and the original recognition failure times threshold is 3 times, which is adjusted to 2 times at this time, so as to start the artificial recognition process earlier.

[0077] The artificial processing module further includes an error identification strategy unit for identifying the cause of failure by image acquisition condition judgment. It collects the illumination parameters, saturation and color temperature values of the current image and the database image. The illumination parameter can be represented by lux, the saturation refers to the richness of the color of the image, and the color temperature value is represented by Kelvin. Then the environmental difference degree is judged by using the image brightness histogram difference and the principal component contrast. The image brightness histogram difference is obtained by calculating the distance between the brightness histograms of the two images, and the principal component contrast is the difference between the principal components of the images. When the difference degree index exceeds the preset threshold, an auxiliary identification light adjustment suggestion is generated, including parameters such as light supplement direction, light intensity level, light source color temperature, and is synchronized to the corresponding access control terminal control system. For example, the illumination parameter of the current image is 200 lux, and the illumination parameter of the database image is 500 lux, the illumination difference is large, and the difference degree index exceeds the threshold, so the light supplement direction of the access control terminal is suggested to be adjusted to front light supplement, the light intensity level is adjusted to level 3 (assuming that the light intensity level is divided into 1 to 5 levels, level 3 is medium intensity), and the light source color temperature is adjusted to 5500 Kelvin. The auxiliary identification light adjustment suggestion is uploaded to the blockchain as an auxiliary identification suggestion data packet to ensure that the adjustment record is traceable.

[0078] The access management platform also includes an abnormality identification and security linkage module, which includes an identity association monitoring unit, a behavior pattern identification subunit, and a security warning pushing unit. The identity association monitoring unit is used to confirm whether the legal identity is successfully matched in each face recognition comparison process. If it cannot be matched, an unbound identity code is added to the identification failure record. The unbound identity code is used to identify that the person has not registered identity information in the system. The behavior pattern identification subunit is used to construct an abnormal behavior pattern vector based on continuous identification failure at the same terminal, attempt to enter and exit at night, and multi-target identification failure with the same person, etc. The abnormal behavior pattern vector is a vector composed of multiple behavior characteristic parameters, such as the number of continuous identification failures, attempt to enter and exit time, and the number of people in the same group. Then it is compared with the existing behavior model on the chain to determine whether it belongs to an abnormal behavior. The security warning pushing unit is used to submit the event as an abnormal event data block to the blockchain network and synchronously push it to the security personnel terminal when the abnormal pattern is triggered. The warning content includes the terminal location such as the access control of the east gate of the community, the image snapshot, i.e. the collected face image, the behavior description such as 3 times of identification failure at 11 pm at night, and the recommended response operation such as going to the scene to check. This module supports linkage calling with the abnormal behavior event library on the chain to form community-level abnormal type trend statistics and automatic early warning rule updating. For example, through statistics, it is found that there are more abnormal events at the west gate of the community on Friday night every week, and the system will automatically update the early warning rules to strengthen the monitoring of the west gate on Friday night.

[0079] The access control management platform further comprises a visitor permission management module, which comprises a visitor registration subunit, a visitor entry subunit and an authorization generation subunit. The visitor registration subunit is used to receive the identity information, face image and access reservation time of the visitor through a two-dimensional code, a self-service terminal or a property APP, such as the visitor uploading his own ID photo, face image and reserving the access time from 2 pm to 4 pm through the property APP. The visitor entry subunit is used to standardize the above information and incorporate it into the visitor feature library, and set the effective access time and the target household, the effective access time being the visitor's reserved time, and the target household being the household the visitor wants to visit, such as the 502 household in 4 unit of 3 building. The authorization generation subunit is used to generate the corresponding visitor access credential and synchronize it to the blockchain network to ensure the trusted record of the authorization chain, and generate the authorization record chain credential hash identifier for subsequent verification, which is a unique string obtained by processing the authorization record through a hash algorithm. The visitor permission management module updates the visitor access state on the chain, such as the visitor successfully entering the community, the access state will be updated to have entered, and synchronized to the blockchain.

[0080] The system further comprises a dynamic information linkage and display module, which comprises an information identification association unit, an information push processing unit, an interactive display unit and a push synchronization subunit. The information identification association unit is used to read the on-chain identity identifier hash value corresponding to the identification event after the access control identification is successful, and call the corresponding visitor or household archive data on the chain. The identity identifier hash value is a unique identifier obtained by processing the identity information through a hash algorithm, through which the corresponding archive data can be accurately found. The information push processing unit is used to determine whether there is un-read visitor record, express visit, access record change and other push information based on the identification identity, and generate corresponding notification data according to the type, such as identifying a certain household, judging that the household has an un-read express visit information, and generating corresponding notification data. The interactive display unit is used to display the push information in the form of text or voice on the access control screen, the property front desk or the mobile application end, such as displaying the text information "You have an express arrived at the property front desk, please pick it up in time" on the access control screen. The push synchronization subunit is used to submit each push record to the blockchain network in a tamper-proof format to ensure the authenticity and traceability of the push record.

[0081] The blockchain network is constructed based on a consortium chain, and each terminal node has data writing right. The blockchain network adopts a specific mechanism to ensure data consistency and credible record. The node writing rule requires that each writing includes node signature, event type code, structured data digest and timestamp. The node signature is the digital signature of the node, which is used to prove that the data is sent by the node; the event type code is the code that identifies the event type, such as the access control identification success code 001; the structured data digest is the result of digest processing of the written data, which is used to quickly verify the data integrity; the timestamp is the specific time of data writing, accurate to seconds. The smart contract rule library includes identification success contract, identification failure contract, exception handling contract, visitor authorization contract and data update contract. These contracts are pre-written programs that are automatically executed when the corresponding event occurs, such as the identification success contract that is automatically executed after the face recognition is successful to record relevant information. The synchronization control logic adopts a synchronization strategy based on polling and event triggering. The polling means that the node regularly asks other nodes whether there is new data, and the event triggering means that the node is actively notified when there is data writing. The on-chain smart contract compares the version number with the preset synchronization period according to the version number comparison rule and the preset synchronization period, which can be 10 minutes, to ensure that the data of each node is synchronized. The historical audit mechanism makes all face recognition events, access records, and exception responses form multi-dimensional identification fields on the chain, including event type, time, location, and processing result, which facilitates subsequent audit and query.

[0082] The following is an example of an application scenario. Take the security system application of a certain community as an example. The community is installed with the security system based on face recognition, and the blockchain network is composed of the management terminal of the community property and the face recognition access control terminal of each entrance as a node to form a consortium chain. When the community residents move in, they enter their identity information and face image through the identity information management module of the management terminal. These information are stored in the face data and identity database, and are synchronized to other nodes through the first data package. When the resident comes home and walks to the community entrance, the face recognition access control terminal collects the face image, compares it with the face features in the database, and outputs the opening signal after successful comparison. The resident enters the community, and the access control event record is stored in each node through the blockchain network.

[0083] A visitor wants to visit a resident in the community, and registers the identity information and face image in the visitor registration subunit through the property APP, and reserves the visit time from 10:00 to 12:00, and the target resident is 2#1#201. The visitor entry subunit processes these information and puts them into the visitor feature library, and the authorization generation subunit generates the visitor access credential and synchronizes it to the blockchain. When the visitor arrives at the community entrance within the reserved time, the face recognition access control terminal collects the face image, compares it with the visitor feature library, and opens the door after successful comparison, and the visitor access state is updated and synchronized to the blockchain.

[0084] If an outsider tries to enter the community and the face recognition fails for three consecutive times, the anomaly handling module starts the manual processing flow, the property management personnel check the relevant images through the manual processing module, judge it as an outsider, and the anomaly chaining confirmation subunit writes the processing result into the blockchain. At the same time, the security response module generates an abnormal warning to inform the security personnel to go to the scene to handle it, and the security personnel feedback the processing situation to the system after arriving at the scene, and the relevant records are synchronized to the blockchain.

[0085] In this process, all data interaction and records are stored and synchronized through the blockchain network, ensuring the security and tamper resistance of the data, and the modules work together to achieve efficient security management of the community.

[0086] The above is only a preferred embodiment of the present application, and the protection scope of the present application is not limited to the above-mentioned embodiments only, any technical solution belonging to the idea of the present application shall be within the protection scope of the present application. It should be noted that for ordinary technical personnel in the technical field, some improvements and decorations without departing from the principle of the present application shall be considered as the protection scope of the present application.

Claims

1. A security system based on face recognition, characterized in that, The application relates to a face recognition and access control system. The system comprises: a blockchain network for realizing distributed storage and synchronous verification of face recognition data and access event records, wherein the blockchain network adopts a consortium chain architecture; a plurality of terminal devices, including a face recognition access control terminal and a management terminal, each of which is one of nodes in the blockchain network and has data writing and reading functions; a face data and identity database distributed in part of the terminal nodes, used for storing face feature information and identity information of community residents and visitors, wherein the database is configured with a first data packet for data synchronization, and each database synchronizes the encrypted data content to other nodes through the blockchain network at a preset time interval or in an event triggered manner; an information processing module configured with a second data packet, used for verifying the structural integrity and encryption consistency of the received first data packet, and performing version control and update confirmation through a smart contract; 2. The face recognition based security system according to claim 1, wherein: an access control management platform used for unified scheduling and policy execution of identity data recording, face recognition, abnormal response and security linkage. The access control management platform comprises an identity information management module, an access control monitoring module, an abnormal processing module and a security response module; The identity information management module is used for inputting and managing the identity information and face data of residents and visitors; The access control monitoring module is used for collecting personnel face images and comparing the face features in the database; If the recognition is successful, an opening signal is output, and if the recognition fails, the record is uploaded to the blockchain; The abnormal processing module is used for starting an artificial recognition process when the number of identification failures exceeds a threshold value; 3.The security system based on face recognition according to claim 1, characterized in that: The security response module is used for generating an abnormal alarm and notifying security personnel to handle. The identity information management module comprises: a data input unit used for receiving and format processing the identity information and face images of community residents or visitors; a resident file subunit used for storing the identity information and face feature vectors of long-term residents and associating the identity information with the resident number; a visitor binding subunit used for establishing a binding relationship between the identity information of temporary visitors and the collected face images and generating a temporary access credential; a visitor file subunit used for saving the bound visitor face and identity record and setting an effective access time window; 4. The face recognition based security system according to claim 3, characterized in that: a data chain control unit used for uploading the newly added or changed information in the resident file and the visitor file to the blockchain network through the first data packet to realize node synchronization.

5. The face recognition based security system according to claim 1, wherein: The identity information management module further comprises an information updating unit, which is used for acquiring the face images collected by the access control monitoring module and comparing the difference degree with the historical face data, and if the difference degree exceeds a set threshold value, the image updating is automatically triggered and uploaded to the blockchain network to replace the old data. The abnormal processing module comprises: an interruption detection unit used for judging whether the same face image is collected multiple times within a preset time period and is not matched successfully; an artificial processing module used for receiving abnormal images and video data, performing artificial identity verification, and judging whether the abnormal images and video data are of registered users or misrecognition behaviors. An abnormality uplink confirmation subunit is configured to pack the artificial processing result, the corresponding image data, the operation record and the identity of the auditor into an abnormality processing data block, and write the processing result into a blockchain network through a preset abnormality event uplink contract; The abnormality processing module is further configured with a threshold dynamic adjustment strategy, which comprises dynamically adjusting the threshold of the number of recognition failures in combination with the frequency of historical abnormal event types on the chain.

6. The face recognition based security system according to claim 5, characterized in that: The artificial processing module further comprises an error recognition strategy unit, which is configured to judge the reason for the recognition failure through image acquisition conditions, including: Acquire the illumination parameters, saturation and color temperature values of the current image and the database image, and judge the environmental difference degree by using the image brightness histogram difference and principal component comparison; When the difference degree index exceeds a preset threshold, generate an auxiliary identification light adjustment suggestion, including light supplement direction, light intensity level, light source color temperature and other parameters, and synchronize to the corresponding access control terminal control system; The auxiliary identification light adjustment suggestion is uploaded to the blockchain as auxiliary identification suggestion data packet.

7. The face recognition based security system according to claim 1, wherein: The access management platform further comprises an abnormality recognition and security linkage module, which comprises: An identity association monitoring unit is configured to confirm whether the legal identity is successfully matched in each face recognition comparison process, and add an unbound identity code to the recognition failure record if the identity cannot be matched; A behavior pattern recognition subunit is configured to construct an abnormal behavior pattern vector based on continuous recognition failures at the same terminal, attempts to enter and exit at night, multiple target recognition failures with the same person, and compare it with the existing behavior model on the chain; A security warning pushing unit is configured to submit the event as an abnormal event data block to the blockchain network and synchronously push it to the security personnel terminal when the abnormal pattern is triggered, and the warning content includes the terminal location, image snapshot, behavior description and suggested response operation; The module supports linkage calling with the abnormal behavior event library on the chain to form community-level abnormal type trend statistics and automatic early warning rule updating. 8.The security system based on face recognition of claim 1, wherein: The access management platform further comprises a visitor permission management module, which comprises: A visitor registration subunit is configured to receive the identity information, face image and access reservation time of the visitor through a two-dimensional code, a self-service terminal or a property APP; A visitor input subunit is configured to standardize the above information and input it into the visitor feature library, and set the effective access time and the target resident; An authorization generation subunit is configured to generate a corresponding visitor access credential and synchronize it to the blockchain network to ensure the trusted record of the authorization chain, and generate an authorization record on-chain credential hash identifier for subsequent verification; The visitor permission management module updates the visitor access state on the chain. 9.The security system based on face recognition of claim 1, characterized in that: Further comprising a dynamic information linkage and display module, which comprises: An information recognition association unit is configured to read the on-chain identity hash value corresponding to the identification event after the access control identification is successful, and call the corresponding visitor or resident archive data on the chain; An information pushing processing unit is configured to judge whether there is un-read visitor record, express delivery visit, access record change and other pushing information based on the identification identity, and generate corresponding notification data according to the type. An interactive display unit is configured to display the push information in the form of text and images or voice on the access control screen, property front desk or mobile application end. A push synchronization subunit is configured to submit each push record to a blockchain network in an unalterable format.

10. The face recognition based security system according to claim 1, wherein: The blockchain network is constructed based on a consortium chain, and each terminal node has data writing right. The blockchain network adopts the following mechanisms to ensure data consistency and credible evidence storage: Node writing rule: each writing needs to include node signature, event type code, structured data digest and timestamp; Smart contract rule library: including successful contract identification, failed contract identification, exception handling contract, visitor authorization contract and data update contract; Synchronization control logic: a synchronization strategy based on polling and event triggering is adopted, and the on-chain smart contract is compared according to the version number comparison rule and the preset synchronization period; Historical audit mechanism: all face recognition events, access records, abnormal responses and the like form multi-dimensional identification fields on the chain.

Images