Entrance and exit person certificate verification method based on prison gate control

Abstract

The application relates to the technical field of information auditing, and discloses an entrance and exit person and certificate verification method based on prison gate management and control, aiming to solve the problems of insufficient adaptation of differentiated security requirements and lack of linkage between permission verification and physical control in the prior art. The method comprises the following steps: generating four types of frequency labels of high frequency, medium frequency, low frequency and sporadic based on historical passing data; designing a differentiated verification process containing feature collection, permission association and gate control instruction for different labels, wherein high-frequency personnel are quickly matched through a dynamic white list, medium-frequency personnel are subjected to multi-factor verification, low-frequency personnel are verified by relying on hash mapping of legal documents and face features, and temporary permission time limit verification is activated in sporadic scenarios; finally, the AB gate interlocking mechanism is linked to ensure that one gate is opened while the other gate is closed, and the results are synchronized to the security system to form a closed loop. The application balances efficiency and safety through classified control, strengthens physical isolation and process compliance, and improves the strictness and dynamic adaptability of prison gate management and control.

Description

Technical Field

[0001] This invention relates to the field of information verification technology, and in particular to a method for verifying identity documents at entrances and exits based on the control of prison gates. Background Technology

[0002] In high-security settings such as prisons, access control is a core element of security. Traditional access control systems primarily rely on mechanical locks, passwords, or single biometric identification technologies, posing significant security risks: manual verification is inefficient and prone to misjudgments due to fatigue or forged documents; early biometric systems had limited functionality and could not handle complex scenarios. Furthermore, traditional solutions often employ localized deployments, lacking real-time integration with cloud data, making it difficult to meet the needs of cross-regional identity verification. While biometric technology has been widely applied in security in recent years, most systems only integrate 2-3 modalities and lack sufficient accuracy in recognizing low-quality images. Although the integration of edge computing and cloud computing has improved real-time performance, constructing a comprehensive, three-dimensional prevention and control system integrating people, identification documents, and behavior remains a challenge in prison settings.

[0003] Existing technology announcement number CN120183077A discloses an AB-door biometric identification control system, suitable for single pedestrian access in scenarios such as prisons and computer rooms. The system consists of a data application server, an A / B door biometric visual intercom control host, and an interactive terminal door unit. It uses biometric technology to verify the identity of personnel entering and exiting. If verification fails, the administrator cannot manually open the corresponding door. This existing technology employs a mechanism of multiple units / personnel controlling different doors, combining automatic system verification with manual review for dual protection, ensuring traceable identity records, eliminating favoritism, strengthening access security, and effectively improving the strictness and security of personnel access control in detention facilities. Existing technology announcement number CN108875478B also discloses a method, device, system, and storage medium for identity verification. This method collects facial images of users to be authenticated in real time and caches data for a preset time period. After obtaining the facial image of the ID card, it selects images that meet the requirements from the cache for identity verification. By reducing real-time processing latency through caching mechanisms, efficient and rapid identity verification is achieved, significantly improving verification efficiency and user experience.

[0004] Regarding the above-mentioned and existing related technologies, the inventors believe that the following defects often exist:

[0005] 1. Existing technologies are insufficiently adapted to differentiated security needs. Current solutions employ a uniform verification process, failing to design differentiated rules based on the frequency of entry and exit of prison personnel and the characteristics of different scenarios. For example, police officers with high-frequency access use the standard verification process, leading to low efficiency and impacting daily supervision; while released prisoners with low-frequency access lack mandatory verification linked to legal documents, creating compliance verification loopholes. This one-size-fits-all approach cannot balance the differentiated needs of prison management for efficiency and security, easily creating regulatory blind spots.

[0006] 2. The existing technology lacks linkage between access verification and physical control. Traditional technologies only achieve superficial matching of identity documents and do not deeply integrate with the unique access control system of prisons and detention centers, leading to the risk of identity matching but access not matching. At the same time, it cannot meet the physical security requirement that the other door must be locked when one door of the prison's A and B doors is open. The verification results and door control status lack a real-time linkage mechanism, which poses a risk of tailgating and intrusion due to abnormal door control equipment or human error. It is difficult to build a complete security closed loop from verification to access control to door control. Summary of the Invention

[0007] The technical problem to be solved by this invention is that the existing technology has the disadvantages of low efficiency and insufficient compliance. To this end, we propose an entrance and exit verification method based on the control of prison gates.

[0008] To achieve the above objectives, this application adopts the following technical solution: a method for verifying identity at prison gates, comprising the following steps: S1: Based on historical passage data of personnel at prison gates, four types of frequency tags are generated: high-frequency, medium-frequency, low-frequency, and occasional. The historical passage data includes frequency, time period, A / B gate passage routes, and prison-specific security level elements. The frequency tags support event-driven updates; S2: Differentiated verification processes are designed for different frequency tags. The processes include closed-loop logic of feature collection, permission association, and gate control command generation, and support multi-target synchronous verification; S3: For high-frequency personnel, dynamic verification is performed by binding them to their job permissions. After the whitelist completes rapid matching and synchronously verifies the compliance of the time period, an AB gate release instruction is generated; S4: For medium-frequency personnel, the business appointment system is linked to perform multi-factor verification of document information, appointment information, and facial features; S5: For low-frequency personnel, a unique mapping relationship between key information in legal documents and facial features is established through a hash algorithm, and verification is completed; S6: For occasional scenarios, temporary permissions containing access time and gate access are activated, and timeliness verification is performed; S7: The verification results from S3 to S6 are linked to the AB gate interlocking mechanism to ensure that when one gate is open, the other gate remains locked, and all verification results are synchronized to the prison security system to form a closed-loop control system.

[0009] Preferably, the frequency tag generation step includes: collecting entry and exit records at the prison gate, extracting core features, and constructing a feature vector. ,in: The frequency parameter is normalized to [0,1]. This is a time period compliance parameter, with a quantized value range of [0,1]. These are fixed parameters for the AB gate route, with quantized values ​​ranging from [0,1]. The risk coefficient parameter for associated personnel is quantified within the range of [0,1]. Frequency labels are generated by weighting the feature vector using a security level weighting algorithm. This algorithm includes dynamically generating a weight matrix based on the prison security level parameter S. The weights satisfy ,and Follow The changes are correlated; the overall score is calculated. By setting a preset threshold range Mapped to high-frequency, medium-frequency, low-frequency, or occasional tags.

[0010] Preferably, the dynamic whitelist includes facial feature vectors of high-frequency individuals. The job authority parameters and time range parameters are dynamically updated using an incremental learning algorithm. The incremental learning algorithm includes: when adding a new employee, calculating their feature vector. The similarity d with the feature vectors in the whitelist ,when hour, For the similarity threshold parameter, Add to whitelist; when an employee leaves or is transferred, their permission parameters are set according to... Attenuation, when In practice, feature vectors are retained as historical archives; periodically, the mean offset between newly added features and historical features is used as a basis for further analysis. The model can be adaptively fine-tuned without retraining on all historical data.

[0011] Preferably, in the verification process for mid-frequency personnel, the business appointment information and the verification result are forcibly associated. After obtaining the personnel's ID information and appointment number, the consistency of the ID, appointment information and facial features is verified through a two-way matching algorithm. If any link does not match, the process is prohibited from being allowed to proceed. Temporary task work order association is also supported.

[0012] Preferably, in the verification process for low-frequency personnel, the key information of legal documents is extracted and adapted to prison-specific documents, including release certificates, transfer notices, and medical approval forms. The personnel number, release date, and escort police officer information are extracted using a keyword positioning algorithm, and a hash mapping relationship is established with facial features.

[0013] Preferably, the AB gate linkage control logic includes: the verification result of a single person must simultaneously satisfy the requirement that one gate is verified and the other gate is locked; in a mixed permission scenario, after all associated personnel have been verified and the escort task work order and document information are matched, an AB gate release instruction is generated; if the verification is passed but the other gate is detected to be unlocked or the information does not match, the release is automatically delayed and an abnormal gate status warning is triggered.

[0014] Preferably, in the verification process for the occasional scenario, the temporary permission follows a closed-loop rule of application, approval, expiration and retrospection. The temporary permission includes the applicant, the time period, the designated gate permission and related task information, and generates an electronic pass code with an expiration time. The permission will automatically expire and be revoked after the preset time period. In emergency scenarios, high-frequency personnel can obtain proxy verification rights through dual approval, pass first, and then complete the document verification records of low-frequency personnel within the specified time.

[0015] A prison gate identity verification system includes a frequency classification module that generates frequency tags based on prison access data and a security policy library, supporting event-driven updates and task pre-marking. The security policy library stores prison-specific policies such as high-frequency time periods for positions and rules for detainees leaving the prison. A differentiated verification module includes a high-frequency whitelist matching unit, a mid-frequency appointment association unit, a low-frequency document binding unit, and an occasional permission control unit, supporting multi-target parallel verification and cross-tag result linkage. A gate control linkage module generates A and B door opening / closing commands based on the verification results, monitors door status, and triggers abnormal warnings.

[0016] Preferably, the differentiated verification module has a built-in mixed scenario processing subunit, which is used to synchronously process the parallel verification of personnel with different frequencies in mixed permission scenarios and output a combined release result.

[0017] Preferably, the door control linkage module has an abnormal circuit breaker and backtracking mechanism. When multiple consecutive verifications fail, it automatically switches to manual review mode. When the physical state of the door does not match the command, it immediately triggers the full door access control to lock and simultaneously alarms the duty room. In emergency proxy verification scenarios, it automatically records the information to be supplemented and triggers a timeout reminder.

[0018] The technical effects and advantages of this invention are as follows:

[0019] This invention establishes four frequency categories—high-frequency, medium-frequency, low-frequency, and occasional—and designs differentiated verification processes for different personnel. The core of this approach lies in categorized management and on-demand adaptation. High-frequency personnel utilize a dynamic whitelist for rapid matching, reducing redundant operations in daily access. Medium-frequency personnel undergo multi-factor verification using documents, appointment information, and facial features to ensure strict matching of permissions with business scenarios. Low-frequency personnel undergo compliance verification through a unique mapping of key information in legal documents to facial features. Occasional scenarios are addressed through time-sensitive control of temporary permissions, balancing emergency response and security auditing. This mechanism solves the problem of balancing efficiency and security in the traditional one-size-fits-all approach, ensuring efficient access for high-frequency personnel while strengthening security for low-frequency and special scenarios through multi-level verification rules.

[0020] In this invention, verification results are deeply integrated with the A / B door interlocking mechanism. By mandating that verification at one door passes while the other door remains locked, the unique physical isolation requirements of prisons are ensured. In mixed access scenarios, a release command is only generated after all associated personnel have passed verification and task information matches, avoiding security risks caused by single verification vulnerabilities. Simultaneously, verification results are synchronized to the prison security system in real time, forming a complete closed loop from feature collection to access verification, door control commands, and status feedback. This solves the problem of disconnect between verification and door control in traditional technologies, significantly reducing the risks of tailgating and intrusion due to equipment malfunctions or human error, and improving the tightness of physical security in prisons.

[0021] This invention achieves system adaptability to scenarios such as personnel changes and permission adjustments through event-driven frequency tag updates, a dynamic whitelist optimized by incremental learning algorithms, and anomaly circuit breaking and backtracking mechanisms. Frequency tags can be updated in real time with job adjustments and task changes, avoiding permission redundancy or gaps; the whitelist dynamically incorporates new personnel and reduces the permissions of departing personnel through incremental learning, eliminating the need for a full model reconstruction; in abnormal situations, it automatically switches to manual review and triggers alarms, while also supporting permission backtracking and completion in emergency scenarios. This dynamic adaptability ensures the system's continuous and stable operation in the complex environment of personnel mobility and permission iteration in prisons and detention centers, reducing manual intervention costs and improving long-term management effectiveness. Attached Figure Description

[0022] The disclosure of this invention is illustrated with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. In the drawings, the same reference numerals are used to refer to the same parts:

[0023] Figure 1 This is a schematic diagram of the workflow structure of the present invention; Detailed Implementation

[0024] It is readily understood that, based on the technical solution of this invention, those skilled in the art can propose various interchangeable structural methods and implementations without altering the essential spirit of the invention. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solution of this invention.

[0025] Reference Figure 1 As shown, the present invention provides a prison gate identity verification system: it adopts a two-level architecture design of edge nodes and cloud hub, and realizes closed-loop control of the entire process from data collection to permission verification and then to gate control execution through the collaboration of distributed computing and centralized management.

[0026] The system's hardware entities are functionally divided into a front-end perception layer, an edge processing layer, a cloud service layer, and an execution control layer. Each layer interacts with data through an encrypted communication protocol, specifically including:

[0027] The front-end sensing layer is deployed in the physical passageways of the prison's A and B gates and surrounding areas. The front-end sensing layer includes the following devices:

[0028] High-definition face capture equipment: It adopts a wide dynamic range camera and has backlight compensation and occlusion detection functions. It is used to capture facial images of people passing through and supports outputting usable features in complex environments such as strong light, shadow, and mask obstruction commonly seen at prison gates.

[0029] Document information collection equipment: integrates OCR scanner and RFID card reader, supports information extraction of documents such as second-generation ID card, lawyer's certificate, temporary pass, etc., and reads the data of the built-in chip of the document for authenticity verification;

[0030] Door status monitoring equipment: Electromagnetic lock status sensors and infrared beam detectors are installed on doors A and B respectively to detect the opening and closing status of the doors. The sampling frequency is 100ms / time to ensure the real-time nature of the door status data.

[0031] Interactive terminal: Includes a touch screen and sound and light prompts, used to display verification progress and anomaly alerts, and to receive manual intervention instructions from the administrator in emergency scenarios.

[0032] The edge processing layer is centered around edge computing terminals deployed in the AB gate control room. It uses industrial-grade embedded processors with a computing power of no less than 8 TOPS. The main tasks of the edge processing layer are:

[0033] It receives raw data from the front-end perception layer in real time, including face images, ID information, and door status signals, and performs preprocessing, such as face image denoising and ID information format standardization.

[0034] Algorithm modules with low latency requirements, such as high-frequency personnel whitelist matching and real-time A / B gate status comparison, ensure localized and rapid response of core processes;

[0035] Data is filtered according to preset rules to reduce the pressure on cloud transmission.

[0036] The cloud service layer consists of a server cluster deployed in the prison's data center, including application servers, database servers, and algorithm servers. Its main functions include:

[0037] Store all historical data, such as personnel access records, whitelist feature databases, and security policy parameters.

[0038] Run complex algorithm modules such as frequency tag generation, low-frequency personnel document hash mapping, and mixed scenario permission combination judgment, and rely on cloud computing power to support deep computing of multi-dimensional data;

[0039] Achieve cross-system collaboration: Establish data synchronization links with existing prison systems, including police personnel systems, visitation appointment systems, and document management systems, through standardized API interfaces to ensure the consistency of basic data.

[0040] The execution control layer includes AB gate controllers, electromagnetic lock drivers, and alarm devices. It receives gate control commands from the edge processing layer or cloud service layer and drives physical equipment to move by outputting control signals through relays. At the same time, it feeds back the execution results of the equipment to the gate control linkage module to form a closed-loop control.

[0041] The system's software functionality is achieved through the collaborative efforts of three core modules: a frequency classification module, a differential verification module, and a gating linkage module. Each module adopts a microservice architecture, enabling independent deployment and elastic scaling. Asynchronous communication between modules is achieved through message queues, preventing single points of failure from affecting the overall process.

[0042] Specifically, the frequency classification module is deployed on the algorithm server in the cloud service layer. Its core function is to generate four types of frequency tags based on historical traffic data. Its operational logic includes:

[0043] Data input: Read the historical access records of personnel from the database server, including entry and exit time, access gate A / B number, associated personnel information, etc., and synchronize the real-time security level from the prison security management system;

[0044] Feature processing: The built-in feature extraction engine is called to calculate feature parameters such as entry and exit frequency and time period compliance from the raw data, and then normalizes them according to preset rules.

[0045] Tag generation: Run the security level weighting algorithm, dynamically adjust the weight ratio of each feature according to the real-time security level, calculate the comprehensive score and map it to the corresponding frequency tag, and then push the tag to the differential verification module through the message queue.

[0046] Specifically, the differentiated verification module is deployed across the edge processing layer and the cloud service layer. The high-frequency whitelist matching unit and the multi-target parallel processing unit are deployed on the edge terminal to ensure real-time performance, while the mid-frequency appointment association unit, the low-frequency document binding unit, and the occasional permission control unit are deployed in the cloud. Its core logic is to invoke the corresponding verification process based on the received frequency tag.

[0047] Input reception: Obtain real-time facial and ID data from the edge terminal, and obtain personnel tags from the frequency classification module;

[0048] Process scheduling: Trigger the corresponding verification unit according to the tag type, such as triggering the whitelist matching unit for high-frequency tags, and coordinate multiple units to work in parallel, such as starting high-frequency and low-frequency verification processes at the same time in a mixed scenario;

[0049] Output results: Package the verification pass / fail results and related data, such as document number and appointment number, and send them to the access control linkage module.

[0050] Specifically, the gating linkage module deploys some functions at both the edge and the cloud: the edge is responsible for the real-time execution and status monitoring of gating commands, while the cloud is responsible for the centralized determination of anomalies. Its core logic includes:

[0051] Command generation: After receiving the pass result from the differential verification module, first query the real-time status of the AB gate uploaded by the edge terminal. If the interlocking conditions are met, generate a gate opening command; otherwise, trigger an abnormal warning.

[0052] Anomaly Handling: Continuously monitor the verification process and gate control execution results. When anomalies such as continuous verification failures or discrepancies between the gate status and the command occur, trigger an alarm according to the preset strategy, such as sending an audible and visual alarm to the duty room or freezing the corresponding channel.

[0053] Log recording: All gating operations and abnormal events are sorted by timestamp and written to the database server to form an audit log, supporting subsequent traceability.

[0054] To achieve data sharing and business collaboration, the system needs to interface with the three existing core systems of the prison. The interface adopts a loosely coupled model, using middleware to achieve data conversion and format adaptation, avoiding direct modification of the original system.

[0055] Integration with the police personnel system: Information such as police officers' job adjustments and departures is synchronized through hourly scheduled tasks, which is used to dynamically update the permission parameters of the whitelist of high-frequency personnel;

[0056] Integration with the meeting appointment system: An event-triggered mechanism is adopted. When a new meeting appointment is generated or canceled, the system automatically receives data such as appointment number and identification information to support multi-factor verification of mid-frequency personnel.

[0057] Integration with the legal document management system: Scanned copies of documents such as release certificates and medical approval forms are read in real time via the OCR interface. Key fields are extracted and hashed with personnel characteristics to establish a hash mapping for use in low-frequency verification processes.

[0058] Based on the aforementioned prison gate identity verification system, this embodiment also provides an entrance and exit verification method based on gate control retrieval, including the following steps:

[0059] S1, the process of generating frequency tags.

[0060] The core of this step is to analyze the historical access data of prison personnel through the frequency classification module to generate four types of frequency tags: high frequency, medium frequency, low frequency, and occasional. This provides a classification basis for subsequent differentiated verification. This is accomplished collaboratively by the data acquisition subunit, feature extraction subunit, and tag calculation subunit of the frequency classification module, as specifically implemented as follows:

[0061] S11. Collect historical passage data and perform preprocessing.

[0062] The data acquisition subunit of the frequency classification module first obtains the raw data required for generating tags from the prison multi-source system, specifically including:

[0063] Basic access data: Collected in real time through interfaces with the AB gate control system and video surveillance system, covering a unique identifier for each person's passage, including police officer's ID, inmate's ID, visitor registration ID, etc.; it also includes fields such as entry and exit time, AB gate number, and direction of passage.

[0064] Related scenario data: Synchronized from the prison business management system, it includes the relationship between escorting police officers and escorted persons in the escort scenario, which is bound by the escort task work order number; it also includes the reason for the escort, such as daily duty, lawyer visit, release from prison, etc., which are generated by the appointment system or manually entered.

[0065] Security level data: obtained from the prison security command platform. The real-time security level parameter S is divided into 1-5 levels according to the daily management and temporary warning status of the prison. Level 1 corresponds to normal control and level 5 corresponds to emergency status. The security level data also includes regional risk coefficients and temporary warning factors. The risk coefficient of the core prison area is 1.0, the risk coefficient of the outer office area is 0.6, the temporary warning factor is 0 in normal circumstances and 1 in emergencies.

[0066] After preprocessing, the collected raw data enters the feature extraction stage. The preprocessing operations include: removing empty records caused by accidental gate triggering and redundant data uploaded repeatedly; converting the time formats of different systems into a standard format; and establishing an association index for multiple passage records of the same person through the unique identifier of the person and the passage timestamp to ensure data correlation and consistency.

[0067] S22. Extraction and quantization of feature vectors.

[0068] The feature extraction subunit calculates the feature vector as described in claim 2 based on the preprocessed historical traffic data. The extraction logic for each feature parameter is as follows:

[0069] Inbound / outbound frequency parameters This is a normalized value of the total number of passages by the target personnel within a preset period of 30 days, which can be dynamically adjusted through the security policy library. Its calculation method is as follows:

[0070] ;

[0071] in This represents the actual number of passages. This represents the maximum number of passages for the same type of personnel within a given period. The value range is [0,1]. For example, if the maximum number of passes per cycle for a prison police officer is 100, and a police officer passes through 60 times in 30 days, then... It is 0.6.

[0072] Time period compliance parameters This represents the percentage of times a target person passes through a system within a preset compliance period out of the total number of passes. The preset compliance period is set by the security policy library according to personnel category, and its value ranges from [0,1]. For example, if a lawyer passes through a system within a compliance period 8 out of 10 times, then... It is 0.8.

[0073] AB gate route fixed parameters This represents the percentage of times the target personnel use the most frequently used A / B gate route out of the total number of times they pass through, with a value ranging from [0,1]. For example, if a police officer enters through gate A and exits through gate B 90% of the time, then... It is 0.9.

[0074] Related personnel risk coefficient parameters For scenarios involving accompanying persons, the calculation formula is as follows:

[0075] ;

[0076] in The highest risk level is for associated personnel. The risk level for associated personnel is obtained from the security policy database and is divided into levels 1-5. The value range is [0,1]. For example, when escorting detainees with a risk level of 3, then... It is 0.4.

[0077] S13. Tag calculation based on security level weight algorithm.

[0078] The label calculation subunit converts the feature vector into a frequency label using the security level weighting algorithm described in claim 2. The specific process is as follows:

[0079] S131, Weight Matrix Dynamic generation.

[0080] Based on real-time security level parameters Adjust the weight ratio of each feature, and satisfy the following conditions: The specific weight adjustment rule is as follows: when Elevating to a state of emergency, time period compliance weighting Risk coefficient weight of related personnel Increase the weight of entry and exit frequency Reduce to strengthen control over compliance and associated risks during specific time periods; when Reduce to normal levels, frequency of entry and exit weight Route fixedness weight Increase the size to meet the high-frequency travel needs of people.

[0081] S132, Comprehensive score calculation and label mapping.

[0082] The overall score is calculated using the formula. The calculation takes values ​​in the range [0,1], and then maps them to four categories of labels using preset threshold intervals in the security policy library:

[0083] High-frequency tag correspondence For example, police officers who frequently pass through during fixed time periods each day;

[0084] Intermediate frequency label corresponding ), such as a lawyer with whom you meet regularly each week;

[0085] Low frequency tag correspondence For example, detainees who receive medical treatment twice a month;

[0086] Occasional tag correspondence For example, emergency repair personnel entering the area for the first time.

[0087] S14, Event-driven dynamic label updates.

[0088] To ensure real-time matching between frequency tags and personnel's actual status, the frequency classification module supports an event-driven update mechanism: when the prison police personnel system triggers events such as job adjustments or permission changes, such as a police officer being transferred from a high-frequency access post to a low-frequency patrol post, the system receives the event signal through the interface, immediately recalculates the personnel's feature vector and comprehensive score, and completes the tag update within 10 seconds; when the prison business system generates temporary tasks, the module pre-marks the associated personnel with occasional tags, and automatically restores the original tags after the task is completed, ensuring the adaptability of the verification process in temporary scenarios.

[0089] S2. Design a differentiated verification process.

[0090] Through a differentiated verification module, corresponding verification processes are designed for the four types of frequency tags generated in S1: high-frequency, medium-frequency, low-frequency, and occasional. Each process follows a closed-loop logic from feature collection to permission association to gating command generation, and supports multi-target synchronous verification to ensure a balance between control requirements and efficiency in different scenarios.

[0091] Although the four verification processes are designed for different frequency labels, they all include three core steps:

[0092] Feature collection stage: It is necessary to obtain basic information related to personnel identity and scene, including but not limited to facial biometrics, identification information, passage time and target channel information.

[0093] The permission association process involves matching the collected features with preset permission rules. These rules originate from a security policy library and business systems, including visitation permissions in the appointment system and access permissions in the document management system. The matching logic must verify feature validity and permission suitability. Feature validity verifies whether the facial features are clear and whether the identification document is valid. Permission suitability verifies whether the access period is within the permitted scope.

[0094] In the gate control instruction generation process: clear instructions must be output based on the permission association results: when the verification is successful, a pending access instruction is generated, which is triggered to open after the gate control linkage module confirms the status of gates A and B; when the verification fails, a prohibition instruction is generated, and the reason is fed back through the interactive terminal, such as mismatched document information or expired permissions.

[0095] It should be noted that for mixed permission scenarios such as police officers escorting detainees or multiple people traveling together, the differentiated verification module has a built-in mixed scenario processing subunit that supports parallel verification of multiple targets.

[0096] When the system detects multiple associated individuals, the mixed-scene processing subunit automatically identifies the frequency tags of each individual and invokes the corresponding verification processes in parallel. For example, police officers follow the high-frequency process, while detainees follow the low-frequency process. Each process independently performs feature collection and permission association, and after generating its own verification results, the subunit summarizes and determines the results through parallel logic: if all targets pass verification, it outputs a combined pass result; if any target fails, it outputs a combined failure result and clearly identifies the failed step.

[0097] For example, in the escort scenario, it is necessary to simultaneously complete the high-frequency whitelist matching of the escorting police officers and the low-frequency document verification of the escorted persons, and the escort task work order information associated with the two must be consistent in order to be judged as a combination pass, so as to ensure full-element control in the mixed scenario.

[0098] The differentiated verification module achieves precise adaptation between the process and the labels through a modular design. The functions of each unit are as follows:

[0099] The high-frequency whitelist matching unit corresponds to the high-frequency tag process. Its core function is to maintain a dynamic whitelist, store the facial feature vectors, job permissions, and time ranges of high-frequency personnel, and realize rapid feature matching and time period compliance verification.

[0100] The mid-frequency appointment association unit corresponds to the mid-frequency tag process. Its core function is to link with the business appointment system, extract appointment information such as appointment number and meeting person, and perform multi-factor comparison with document information and facial features.

[0101] The low-frequency document binding unit corresponds to the low-frequency tagging process. Its core function is to parse prison-specific documents, such as release certificates and transfer notices, extract key information, and establish a unique mapping relationship with facial features for verification.

[0102] The occasional access control unit corresponds to the occasional tag process. Its core function is to manage the entire lifecycle of temporary access, including application, approval, time limit control, expiration and revoke, and generate an electronic pass code with an expiration date for verification.

[0103] S3, Execution of the high-frequency personnel verification process.

[0104] This process relies on a dynamic whitelist that is bound to job permissions to achieve feature matching, and generates an AB gate release instruction after simultaneously verifying the compliance of the time period.

[0105] Dynamic whitelists are the core basis for verifying frequently used personnel, and their storage structure must meet the requirements of fast retrieval and permission association, specifically including:

[0106] Face feature vector: using d-dimensional real vectors Store the facial biometrics of high-frequency personnel. The feature dimension d is set to 128 or 256 dimensions according to the recognition accuracy requirements. The vector is extracted through a deep neural network and normalized to ensure consistency of cross-device matching.

[0107] Job permission parameters: include the personnel's job position, the A and B access gates, the level of permission effectiveness, etc., which are synchronized in real time with the prison police personnel system;

[0108] Time period range parameter: Records the time period rules for permission, distinguished by weekdays and holidays, and supports permission extension for special time periods, such as during duty hours, which are dynamically configured by the security policy library;

[0109] The whitelist data is stored in a high-speed cache at the edge processing layer, while a backup is maintained at the cloud service layer to ensure low latency and data security for local verification.

[0110] Dynamic whitelists are updated in real time through incremental learning algorithms, without requiring a full model reconstruction. Specifically, this includes:

[0111] New personnel inclusion: When new high-frequency personnel are added to prisons and detention centers, the system automatically collects their facial features to generate vectors. Calculate the Euclidean distance d between this vector and existing feature vectors in the whitelist. If the distance is greater than the preset similarity threshold, then... The associated job permissions and time period parameters are written into the whitelist. The whole process is completed within 5 minutes and does not affect the matching efficiency of the existing whitelist.

[0112] Handling of permissions for departing or transferred employees: When an employee leaves or is transferred to a less frequently used position, their permission parameters are adjusted according to an exponential decay formula. Dynamic adjustment, among which These are the initial permission values. This is the attenuation coefficient, set according to the importance of the position, such as 0.05 / day. This represents the number of days since the employee left the company or was reassigned. When the permission value drops below the freeze threshold, the facial feature vector is retained as a historical archive, but it will no longer participate in real-time matching.

[0113] Model adaptive fine-tuning: The system adjusts the mean shift between newly added features and historical features every 7 days. The matching model is fine-tuned by adjusting the weight distribution of feature vectors to adapt to natural changes in facial features, such as aging and hairstyle changes, to ensure long-term matching accuracy. The fine-tuning process is automatically executed during off-peak hours without interrupting normal verification.

[0114] Meanwhile, verification of high-frequency personnel must simultaneously meet the dual conditions of feature matching and time period compliance: the system obtains the current time in real time and compares it with the time period range parameters of the personnel in the whitelist to determine whether they are within the permitted access period; if they are in the regular time period, the verification is passed directly; if they are in special time periods such as overtime or emergency duty, a temporary authorization code needs to be verified to ensure the controllability of permission expansion; if the time period verification fails, the system prompts through the interactive terminal that there is no access permission for the current time period and simultaneously records the attempted access information to the audit log for subsequent traceability.

[0115] After high-frequency personnel verification is passed, the high-frequency whitelist matching unit outputs a pending release instruction to the gate control linkage module. The instruction includes the personnel's unique identifier, the target channel, and the verification pass time. After the gate control linkage module confirms that the other door is locked, it triggers the target channel opening instruction. The entire process response time does not exceed 500 milliseconds, meeting the needs of high-frequency personnel for rapid passage.

[0116] S4. Execution of the mid-frequency personnel verification process.

[0117] During verification by mid-frequency personnel, the system first completes the collection and synchronization of two types of core information:

[0118] The document information is obtained through the front-end document collection device, including fields such as document type, document number, holder's name, and photo. At the same time, the encrypted information built into the document is read through the RFID chip to verify its authenticity and eliminate the risk of counterfeit documents.

[0119] Appointment information is synchronized in real time from the business appointment system, including appointment number, appointment person information, meeting person, appointment time, designated access channel, etc. The synchronization process is completed through an encrypted API interface to ensure the security of data transmission.

[0120] The system will initially associate the collected document information with the synchronized appointment information, and establish a mapping between the document number and the appointment number to provide basic data for the subsequent two-way matching algorithm.

[0121] The bidirectional matching algorithm is the core of the intermediate frequency verification. It ensures information consistency through three layers of verification. The specific steps are as follows:

[0122] The first layer of verification compares the identification information with the appointment database. The Levenshtein edit distance algorithm is used to calculate the similarity between the identification information and the appointment database records. The formula is:

[0123] ;

[0124] in, This is a key string in the document. The string is the corresponding field in the reservation database. Let be the edit distance between the two strings. Set the matching threshold for the string length. =1.0, meaning a complete match is required; otherwise, it is considered a mismatch.

[0125] The second layer of verification involves comparing facial features with the ID photo. A cosine similarity algorithm is used to calculate the matching degree between real-time facial features and ID photo features; the formula is:

[0126] ;

[0127] in, This is a 128-dimensional facial feature vector acquired in real time. The 128-dimensional feature vector extracted from the photo embedded in the ID card. Represents the L2 norm and sets the matching threshold. =0.85, when If the face matches, it is determined that the face does not match; otherwise, it is determined that the face does not match.

[0128] The third layer of verification is the validity check of the reservation status. First, a time window check is performed, calculating the current time. With appointment time slot The deviation is expressed by the formula:

[0129] ;

[0130] Set allowable deviation threshold =30min, when The time is determined to be valid.

[0131] Next, a status verification is performed, and the reservation status identifier is obtained through the reservation system interface. , =0 indicates invalid. =1 indicates that it is valid. =2 indicates that the reservation has been cancelled. A reservation is considered valid if and only if both the time and status are valid.

[0132] If all three layers of verification pass, the verification is considered successful; if any step fails to match, the process is immediately terminated and an error message is triggered.

[0133] For unscheduled business transactions, such as urgent meetings and official business, the system supports the associated verification of temporary task work orders. The specific process is as follows:

[0134] Temporary task work orders are generated by the prison's internal business system and include work order number, task type, applicant, visitor information, temporary access period, etc. After generation, they take effect through the authorization approval link.

[0135] When verifying, personnel can manually enter a temporary work order number through the interactive terminal. After the system calls the work order interface to obtain the work order information, it will be included as temporary appointment information in the matching process, replacing the regular appointment information to participate in the above three-layer verification. The verification logic is the same as that of regular appointments.

[0136] Temporary work orders are valid for 24 hours by default and will automatically expire after that time to ensure the controllability of temporary permissions.

[0137] If any mismatch occurs during the verification process of medium frequency personnel, the system will immediately execute a prohibition and release operation. The specific response includes:

[0138] The interactive terminal displays the reasons for the mismatch in real time, such as the document information not matching the person making the appointment, the facial features not matching, or the appointment having expired, and guides the personnel to correct the information or terminate the passage.

[0139] The system automatically records detailed logs of mismatch events, including time, personnel information, and mismatch details, and synchronizes them to the prison security system for archiving and subsequent auditing.

[0140] If the same person's verification fails three times in a row, the system will trigger an early warning mechanism, notifying the duty room to intervene manually and investigate any abnormal risks.

[0141] S5. Execution of low-frequency personnel verification process.

[0142] Verification of low-frequency personnel is based primarily on legal documents. The document input and parsing process is as follows:

[0143] Specialized documents include release certificates, transfer notices, and medical treatment approval forms. These are entered into the system by prison management personnel in two ways: first, by scanning paper documents into electronic images using a high-definition scanner and uploading them to the document management system; second, by directly connecting to the prison's electronic document platform to synchronize the generated electronic documents.

[0144] The system uses a keyword localization algorithm to perform structured parsing of documents, extracting core fields for different document types: For release certificates, the released person's ID, release date, and the signature of the approving police officer are extracted; for transfer notices, the transferee's ID, the name of the target prison, and the escorting police officer's ID are extracted; for medical treatment approval forms, the patient's ID, treatment date, and accompanying police officer's information are extracted. The keyword localization algorithm identifies field locations using a pre-set document template and combines this with OCR technology to extract text, ensuring no core information is missed.

[0145] To achieve a unique binding between legal documents and personal characteristics, the system establishes a mapping relationship through a hash algorithm, as follows:

[0146] First, the key information extracted from the documents, such as personnel number, release date, and approval information, is concatenated to form a string. ,For example =Personnel ID + Release Date + Approving Police Officer ID.

[0147] Secondly, the low-frequency facial feature vectors of the collected individuals Serialization is performed, and the data is converted into a computable feature string using Base64 encoding. ;

[0148] Finally, the document string With characteristic strings The input string is combined using the concatenation operator. ,Right now Using the SHA-256 hash algorithm to Perform calculations to generate a 256-bit hash value. The SHA-256 algorithm outputs a fixed-length 64-bit hexadecimal string.

[0149] The generated hash value The data is stored in a low-frequency personnel feature database in association with document number and unique personnel identifier, realizing a one-to-one mapping between documents, faces, and hash values.

[0150] The uniqueness of hash mapping is guaranteed by two mechanisms: first, the non-repeatability of key information in documents; and second, the collision resistance of hash algorithms, ensuring that different people or different documents will not generate the same hash value.

[0151] It should be noted that when low-frequency personnel are verified, the system performs matching according to the following steps:

[0152] The first step is to collect real-time facial features. Extract feature vectors and convert them into feature strings. A hash value to be verified is generated using a hash algorithm;

[0153] The second step involves inputting the document number via the interactive terminal. The system then retrieves the key information of the corresponding document from the document management system and generates the original document string. and base hash value ;

[0154] The third step is to compare the hash values ​​to be verified. Compared with the baseline hash value If they match, proceed to validity verification;

[0155] The fourth step, validity verification, includes: whether the release date is the current date or within the allowed time window, whether the escorting police officer's information matches the accompanying personnel, and whether the document has been marked as used.

[0156] If all the above steps are completed, the verification is considered successful; if the hash value does not match or the validity check fails, the verification is considered unsuccessful, and the reason will be displayed on the interactive terminal.

[0157] To strengthen compliance management in low-frequency scenarios, the system has set up multiple safeguards: document entry requires approval from two people, such as correctional officers entering the document and department heads reviewing it, to ensure the authenticity of the document information; once the hash mapping relationship is established, it cannot be modified. If a change is required, a new document must be generated and the mapping must be re-established, and the original mapping will be marked as invalid; after verification, the system automatically updates the document status to "used" and records information such as usage time and access channel, forming a closed-loop audit with the prison security system to prevent the risk of document reuse.

[0158] S6. Execution of the verification process for occasional scenarios.

[0159] The generation of temporary permissions must go through a standardized application and approval process to ensure the compliance of the permission grant. The applicant initiates a temporary permission application through the prison's internal permission management system, fills out an application form, and includes the following core information: the reason for the application, the information of the authorized person, the designated time period for passage, the gates and passages that are allowed to pass through, and the associated task information.

[0160] After the application is submitted, it enters the approval process: In normal scenarios, it requires two levels of approval, such as initial review by the department head and secondary review by the security department. After approval, a temporary permission record is generated and stored in the permission library of the occasional permission control unit. If the approval fails, the system will provide feedback to the applicant on the reason for rejection, such as the time period being outside the allowed range or the lack of related task information, and the application must be resubmitted.

[0161] After the temporary access is granted, the system automatically generates an electronic pass code with an expiration date, which serves as verification credentials for occasional personnel.

[0162] The electronic access code is generated using dynamic encryption and contains three parts: an authorization identifier, a timestamp, and an encryption checksum, ensuring it cannot be forged. The access code is presented in QR code form and can be sent to the authorized person's mobile phone via SMS, or printed and delivered by the applicant.

[0163] During verification, occasional personnel use a front-end scanning device to read the electronic access code. The system decrypts the code, extracts the permission identifier, queries the permission database to obtain the corresponding temporary permission information, and verifies the following: whether the current time is within the permitted access period, whether the access channel matches the designated gate, and whether the authorized person's identification information matches the information provided during the application. If verification is successful, the verification is considered complete; if any information does not match or the access code has expired, the verification is considered unsuccessful.

[0164] For emergencies such as medical emergency and equipment repair that require immediate intervention, the system supports frequently accessed personnel to obtain proxy verification rights.

[0165] Frequently accessed personnel can initiate a proxy verification request via the interactive terminal, submitting an explanation of the emergency and information about the person being represented. The system automatically triggers a two-person approval process, requiring two on-duty leaders to simultaneously confirm via password or biometrics. Once approved, the frequently accessed personnel gain proxy verification rights for one hour and can lead the occasionally accessed personnel through the system.

[0166] After passage, the system automatically records the proxy information, including the proxy police officer's number, the information of the person being proxied, and the passage time. It then sends a reminder to the proxy police officer to complete the necessary documentation within 24 hours, such as repair work orders or first-aid certificates. Once completed, the system links the documentation information to the proxy record and archives it. Failure to complete the documentation within the deadline triggers a permission warning, freezing the proxy verification rights of the frequently accessed officer until manual review and approval.

[0167] To prevent the abuse of temporary permissions, the system has a strict expiration and revocation mechanism:

[0168] The validity period of the electronic pass code is consistent with the passage period of the temporary permission. After the period expires, the system will automatically mark the pass code as invalid, and the verification will be directly judged as unsuccessful when scanning the code.

[0169] Expired temporary permission records in the permission database are retained for 30 days. After the expiration period, they are automatically archived to the historical database and will no longer be involved in real-time verification.

[0170] If temporary access is revoked during its validity period, the approver can manually trigger a revocation command through the access control system. The system will immediately invalidate the corresponding electronic pass and simultaneously notify the authorized person and the associated police officer.

[0171] The formation of S7, AB gate linkage and control closed loop.

[0172] AB door interlocking is a core requirement of physical security in prisons and detention centers. Its execution logic is achieved through the collaborative action of the door status monitoring subunit and the instruction generation subunit of the door control linkage module.

[0173] When a single person passes through, the system first receives the verification result from the differential verification module, which includes the target access door and the person's unique identifier. The door status monitoring subunit collects the physical status of doors A and B in real time through the electromagnetic lock feedback device and infrared beam sensor. After confirming that the other door is locked, the instruction generation subunit sends an opening instruction to the controller of the target door.

[0174] Once the target door is opened, the system continuously monitors its status until personnel pass through and the door closes. Then, it automatically triggers the verification status of the other door, meaning that subsequent personnel must start verification from the closed door to ensure that the two doors are not open at the same time.

[0175] For scenarios with mixed access permissions, such as police officers escorting detainees or multi-person collaborative operations, the access control linkage module must meet the combined conditions of full verification approval and task information matching, as follows:

[0176] The hybrid scene processing subunit receives the multi-target verification results output by the differentiated verification module, such as the high-frequency verification of police officers passing and the low-frequency verification of detainees passing. First, it verifies that the verification status of all related personnel is passed. Second, it calls the escort task work order information and compares whether the escort police officer number, the escorted person number, the task validity period in the work order are consistent with the personnel information in the verification results. When both are satisfied, it generates a combined release instruction and opens the A and B doors in a preset order. For example, it opens the A door first, and after the personnel enter the middle area and close the A door, it opens the B door.

[0177] If any person fails verification or the task information does not match, the system will immediately terminate the release process, keep both doors locked, and notify the duty room terminal of the mixed scenario verification failure, clearly indicating the failed item.

[0178] When verification passes but the gate status does not meet expectations or there are conflicting information, the gate linkage module handles the situation according to the following strategy:

[0179] If a door is found to be unlocked after verification, the system will automatically delay the release command and issue a warning in the A and B door areas via an audible and visual alarm, and push a door status discrepancy notification to the duty room. If the door status returns to normal during the delay period, the release command will be executed. If the door does not return to normal after the timeout, the verification permission for that channel will be automatically frozen and manual unlocking will be required.

[0180] When information does not match, such as when the escort work order and the document information are inconsistent, the system will terminate the release and record the conflicting fields, and simultaneously trigger the audit log. After verification by the management personnel, the system will unfreeze the information through the manual approval process to ensure that the abnormal situation is traceable and controllable.

[0181] The synchronization and archiving of information throughout the entire process is the core of the closed-loop management system, and its specific implementation is as follows:

[0182] The access control linkage module will synchronize the verification results, access control operation records, abnormal events, etc. to the prison security system in real time;

[0183] The security system stores synchronized information in a structured manner, supports queries by personnel, time, event type, and other dimensions, and generates periodic control reports to provide data support for prison management. Furthermore, all operation records are tamper-proof, meeting compliance audit requirements.

[0184] The technical scope of this invention is not limited to the content described above. Those skilled in the art can make various modifications and variations to the above embodiments without departing from the technical concept of this invention, and all such modifications and variations should fall within the protection scope of this invention.

Claims

1. A method for verifying identity documents at entrances and exits based on prison gate control, characterized in that, Includes the following steps: S1: Based on the historical passage data of personnel at the prison gate, generate four types of frequency tags: high frequency, medium frequency, low frequency and occasional. The historical passage data includes frequency, time period, passage routes of A and B gates and security level elements unique to the prison. The frequency tags support event-driven updates. S2: Differentiated verification processes are designed for different frequency tags. The process includes closed-loop logic of feature collection, permission association and gating instruction generation, and supports multi-target synchronous verification. S3: For high-frequency personnel, a dynamic whitelist bound to job permissions is used to quickly match them, and after verifying the compliance of the time period, an AB gate release instruction is generated. S4: For mid-frequency personnel, link the business appointment system to conduct multi-factor verification of document information, appointment information and facial features; S5: For low-frequency individuals, a unique mapping relationship between key information in legal documents and facial features is established using a hash algorithm, and verification is completed; S6: For occasional scenarios, activate temporary permissions that include access time and gate access, and perform time-sensitive verification; S7: The verification results from S3 to S6 are linked to the AB door interlocking mechanism to ensure that when one door is open, the other door remains locked, and all verification results are synchronized to the prison security system to form a closed-loop control system.

2. The method for verifying identity documents at entrances and exits based on prison gate control as described in claim 1, characterized in that: The frequency tag generation steps include: Collect entry and exit records from the prison gate, extract core features, and construct feature vectors. ,in: The frequency parameter is normalized to [0,1]. This is a time period compliance parameter, with a quantized value range of [0,1]. These are fixed parameters for the AB gate route, with quantized values ​​ranging from [0,1]. This is a risk coefficient parameter for associated personnel, with a quantified value range of [0,1]. The feature vector is weighted using a security level weighting algorithm to generate frequency labels. The security level weighting algorithm includes: A weight matrix is ​​dynamically generated based on the prison security level parameter S. The weights satisfy ,and Follow The changes are related; Calculate the overall score By setting a preset threshold range Mapped to high-frequency, medium-frequency, low-frequency, or occasional tags.

3. The method for verifying identity documents at entrances and exits based on prison gate control as described in claim 1, characterized in that: The dynamic whitelist includes facial feature vectors of frequently used individuals. The job permission parameters and time range parameters are dynamically updated using an incremental learning algorithm, which includes: When adding new personnel, calculate their feature vectors. The similarity d with the feature vectors in the whitelist ,when hour, For the similarity threshold parameter, Add to whitelist; When an employee leaves or is transferred, their permission parameters are adjusted according to... Attenuation, when At that time, the feature vectors are retained as historical archives; Periodically based on the mean shift between new features and historical features The model can be adaptively fine-tuned without retraining on all historical data.

4. The method for verifying identity documents at entrances and exits based on prison gate control as described in claim 1, characterized in that: In the verification process for mid-frequency personnel, the business appointment information and the verification result are forcibly linked. After obtaining the personnel's ID information and appointment number, the consistency of the ID, appointment information and facial features is verified through a two-way matching algorithm. If any link does not match, the process is prohibited from being allowed to proceed. Temporary task work order association is also supported.

5. The method for verifying identity documents at entrances and exits based on prison gate control as described in claim 1, characterized in that: In the verification process for low-frequency personnel, key information from legal documents is extracted and adapted to prison-specific documents, including release certificates, transfer notices, and medical approval forms. The personnel number, release date, and escort police officer information are extracted using a keyword positioning algorithm and a hash mapping relationship is established with facial features.

6. The method for verifying identity documents at entrances and exits based on prison gate control according to claim 1, characterized in that: The AB door linkage control logic includes: the verification result of a single person must simultaneously satisfy the requirement that one door is verified and the other door is locked; In mixed access scenarios, after all associated personnel have been verified and the escort task work order and document information are matched, an AB gate release instruction is generated. If the verification passes but another door is detected to be unlocked or the information does not match, the passage will be automatically delayed and an abnormal door status warning will be triggered.

7. The method for verifying identity documents at entrances and exits based on prison gate control according to claim 1, characterized in that: In the verification process for the occasional scenario, the temporary permission follows a closed-loop rule of application, approval, expiration and retrospection. The temporary permission includes the applicant, the time period, the designated gate permission and related task information, and generates an electronic pass code with an expiration time. The temporary permission will automatically expire and be revoked after the preset time period. In emergency scenarios, high-frequency personnel can obtain the proxy verification right through dual approval, pass first, and then complete the document verification record of low-frequency personnel within the specified time.

8. A prison gate identity verification system implementing the method of any one of claims 1-7, characterized in that, include: The frequency classification module generates frequency tags based on prison access data and a security policy library, and supports event-driven updates and task pre-marking. The security policy library stores prison-specific policies such as high-frequency periods for positions and rules for detainees to leave the prison. The differentiated verification module includes a high-frequency whitelist matching unit, a medium-frequency appointment association unit, a low-frequency document binding unit, and an occasional permission control unit, supporting multi-target parallel verification and cross-tag result linkage; The door control linkage module generates AB door opening and closing commands based on the verification results, monitors the door status, and triggers abnormal warnings.

9. The prison gate identity verification system according to claim 8, characterized in that: The differentiated verification module has a built-in mixed scenario processing subunit, which is used to synchronously process the parallel verification of personnel with different frequencies in mixed permission scenarios and output a combined release result.

10. The prison gate identity verification system according to claim 8, characterized in that: The door control linkage module has an abnormal circuit breaker and backtracking mechanism. When an abnormality is verified multiple times in a row, it automatically switches to manual review mode. When the physical state of the door is detected to be inconsistent with the command, it immediately triggers the full door access control to lock and simultaneously alarms the duty room. In the case of emergency proxy verification, it automatically records the information to be supplemented and triggers the timeout reminder.

Images