A heterogeneous security cooperation control method and device based on face recognition and a medium

Abstract

The application discloses a heterogeneous security cooperation control method and device based on face recognition and a medium, relates to the technical field of intelligent security, and comprises the following steps: continuously detecting frames of face image data to determine security identity data of a user; verifying the security identity data based on a preset trigger rule library to obtain a security operation queue; constructing parameterized execution instructions corresponding to each operation item in the security operation queue through parameter template analysis; obtaining a real-time arming state of a security host, and determining security instruction data through dynamic filling of a key symbol sequence according to the parameterized execution instructions and the real-time arming state; and adjusting the interframe interval of timing parameters corresponding to the security instruction data through instruction response detection to obtain security host control data. The application solves the technical problem that the timing of a key sequence cannot be dynamically adjusted in the face recognition security cooperation control method of the prior art.

Description

Technical Field

[0001] This application relates to the field of intelligent security technology, and in particular to a heterogeneous security collaborative control method, device and medium based on face recognition. Background Technology

[0002] The alarm control panel is the core device of the security system. It is used to connect various detectors and perform operations such as arming and disarming according to user instructions. Its control method is highly dependent on the physical keyboard. Users must enter a numeric password on the dedicated keyboard that comes with the control panel, and then use function keys such as arming, disarming, and bypassing to complete the operation. Facial recognition devices only output switch signals or simple identity signals, and the keyboard protocol compatibility with the corresponding alarm control panel is insufficient.

[0003] In existing technologies, security collaboration typically involves pre-recording fixed key sequences and repeatedly sending them. Due to timing simulation limitations, key values ​​can only be sent at fixed intervals, leading to host rejection or errors. For collaborative methods involving face recognition, a password string is usually output via serial port after recognition, and then received by the host. In face recognition-based security collaborative control, existing technologies suffer from the problem that the timing of key sequences cannot be dynamically adjusted. Therefore, there is an urgent need for a keyless face recognition heterogeneous security collaborative control method. Summary of the Invention

[0004] This application provides a heterogeneous security collaborative control method, device, and medium based on face recognition, which solves the technical problem that the key sequence timing cannot be dynamically adjusted in the existing face recognition security collaborative control method.

[0005] In a first aspect, embodiments of this application provide a heterogeneous security collaborative control method based on face recognition. The method includes: acquiring a user's face image data and performing continuous frame liveness detection on the face image data to determine the user's security identity data; performing verification condition matching on the security identity data based on a preset trigger rule base to obtain a security operation queue, which includes disarming operation items and arming operation items; constructing parameterized execution instructions corresponding to each operation item in the security operation queue through parameter template parsing; acquiring the real-time arming status of the security host and determining security instruction data by dynamically filling a sequence of key symbols based on the parameterized execution instructions and the real-time arming status; and adjusting the frame interval of the timing parameters corresponding to the security instruction data through instruction response detection to obtain security host control data.

[0006] In one implementation of this application, continuous frame liveness detection is performed on facial image data to determine the user's security identity data. Specifically, this includes: locating the facial region in the facial image data and extracting the facial feature vector corresponding to each frame in the facial region; performing dynamic detection on the facial feature vector to determine the facial dynamic frames corresponding to the facial image data; wherein, dynamic detection includes: blink dynamic detection and lip micro-movement detection; performing liveness scoring based on the temporal features of the facial dynamic frames to obtain the true state of the facial image data; wherein, the true state includes: real and fake; the liveness score is determined by accumulating the scores of at least one of the following detections: non-periodic blink interval detection, lip opening / closing degree change detection, and dynamic range detection of the distance between facial feature vectors in continuous frames; when the accumulated score exceeds a preset liveness threshold, it is determined to be real; when it is below the threshold, it is determined to be fake; when the true state is real, the user data corresponding to the facial feature vector is matched to determine the user's security identity data; when the true state is fake, the current recognition request is rejected, an abnormal log is recorded, and the same person is prohibited from re-initiating recognition within a preset cooldown period.

[0007] In one implementation of this application, based on a preset trigger rule base, security identity data is verified by matching the verification conditions to obtain a security operation queue. Specifically, this includes: parsing the user permissions corresponding to the security identity data and reading the current access status of the security host; based on the user permissions and the current access status, matching the corresponding rule parameters in the trigger rule base to determine the operation entries of the security host; wherein, the operation entries include: trigger conditions, host operation templates, and operation priorities; and sorting the operation entries in execution order to obtain the security operation queue.

[0008] In one implementation of this application, parameterized execution instructions corresponding to each operation item in the security operation queue are constructed through parameter template parsing. Specifically, this includes: parsing parameters for each operation item in the security operation queue to determine the placeholder corresponding to the operation item; saving the fixed parameters in the placeholder and replacing the non-fixed parameters in the placeholder with strings to obtain the host execution parameters; encrypting and signing the security identity data, and integrating the encrypted and signed security identity data with the host execution parameters to determine the parameterized execution instructions; wherein, the encryption and signing adopts the HMAC-SHA256 algorithm.

[0009] In one implementation of this application, security instruction data is determined by dynamically filling a sequence of key symbols based on parameterized execution instructions and real-time arming status. Specifically, this includes: acquiring a security keypad template and setting the corresponding signal transmission interval; wherein the security keypad template includes: a regular disarming template, an alarm-extinguishing disarming template, and a zone disarming template; performing virtual keyboard key sequence matching on the real-time arming status to determine the security keypad template corresponding to the real-time arming status; wherein the real-time arming status includes: an alarm status, a multi-zone status, and a default status; and supplementing the parameterized execution instructions into the placeholders corresponding to the security keypad template to obtain the security instruction data.

[0010] In one implementation of this application, virtual keyboard key sequence matching is performed on the real-time arming state to determine the security key template corresponding to the real-time arming state. Specifically, this includes: when the real-time arming state is an alarm state, matching an alarm cancellation / disarming template to obtain a first security key template; when the real-time arming state is a multi-zone state, matching a zone disarming template to obtain a second security key template; obtaining a sequence of historically successful key templates with the same host identifier and operation type as the current host in the historical execution experience database, and using the key arrangement order and inter-frame delay parameters in the historically successful key template sequences as the arrangement and timing reference for the security key template matched in the current real-time arming state to determine the security key template.

[0011] 7. A heterogeneous security collaborative control method based on face recognition according to claim 1, characterized in that, through command response detection, the frame interval of the timing parameters corresponding to the security command data is adjusted to obtain the security host control data, specifically including: setting the initial frame interval corresponding to the security command data, and sending the security command data to the security host via the keyboard bus at the initial frame interval; after sending, switching to receiving mode, receiving the response code returned by the security host by listening to the keyboard bus, and demodulating the response code into the corresponding byte stream to determine the response type; if the response type is successful, recording the currently used frame interval parameter combination as a successful experience, and improving the system. The frame interval parameter combination is selected with weights under the same host and operation type in subsequent operations to determine the first security host control data; in the case of a timing error response, a bidirectional step search strategy is used to adjust the initial frame interval; in the case of a successful response after adjustment, the successful interval value is stored in the historical timing experience database and the optimal interval is updated in a weighted average manner to determine the second security host control data; in the case of a failure response after several adjustments, the host is determined to be currently uncontrollable, a timing anomaly report is generated and the operation entry is skipped; the security host control data is obtained based on the first or second security host control data.

[0012] In one implementation of this application, after adjusting the frame interval of the timing parameters corresponding to the security command data through command response detection to obtain the security host control data, the method further includes: extracting the trigger features of the security host control data and matching the security rules corresponding to the trigger features; wherein, the trigger features include: host identifier, operation type, execution status, and failure reason classification; according to the security rules, determining the operation delay of the host's operation queue through security action sorting, and updating the security host control data based on the operation delay; inserting the updated security host control data into the operation queue to determine the host security operation log.

[0013] Secondly, embodiments of this application also provide a heterogeneous security collaborative control device based on face recognition, characterized in that the device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to implement a heterogeneous security collaborative control method based on face recognition.

[0014] Thirdly, embodiments of this application also provide a non-volatile computer storage medium for heterogeneous security collaborative control based on face recognition, which stores computer-executable instructions, characterized in that the computer-executable instructions, when executed, can realize a heterogeneous security collaborative control method based on face recognition.

[0015] This application provides a heterogeneous security collaborative control method, device, and medium based on face recognition. By extracting features from the face images acquired by the security system, the corresponding personnel's identity data is constructed. Based on the identity data, the operation queue and instruction dynamic processing on the host side, as well as the adaptive mapping of the key sequence, solve the technical problem that the key sequence timing of the face recognition security collaborative control method in the prior art cannot be dynamically adjusted, and realizes the dynamic adjustment of the timing parameters of the heterogeneous security host. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart of a heterogeneous security collaborative control method based on face recognition is provided for an embodiment of this application; Figure 2 This is a schematic diagram of the internal structure of a heterogeneous security collaborative control device based on face recognition, provided as an embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] This application provides a heterogeneous security collaborative control method, device, and medium based on face recognition. By extracting features from the face images acquired by the security system, the corresponding personnel's identity data is constructed. Based on the identity data, the operation queue and instruction dynamic processing on the host side, as well as the adaptive mapping of the key sequence, solve the technical problem that the key sequence timing of the face recognition security collaborative control method in the prior art cannot be dynamically adjusted, and realizes the dynamic adjustment of the timing parameters of the heterogeneous security host.

[0019] The technical solutions proposed in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0020] Figure 1 This document provides a flowchart of a heterogeneous security collaborative control method based on face recognition, as illustrated in an embodiment of this application. Figure 1 As shown in the figure, the heterogeneous security collaborative control method based on face recognition provided in this application embodiment specifically includes the following steps: Step 101: Obtain the user's facial image data and perform continuous frame liveness detection on the facial image data to determine the user's security identity data.

[0021] For example, in order to convert facial images of people in front of security system cameras into machine-readable identity credentials and prevent forgery attacks through photos and videos, this application provides a data foundation for the construction of security operation queues by performing continuous frame liveness detection on facial image data.

[0022] Specifically, continuous frame liveness detection is performed on facial image data to determine the user's security identity data. This includes: locating the facial region in the facial image data and extracting the facial feature vector corresponding to each frame in the facial region; performing dynamic detection on the facial feature vector to determine the facial dynamic frames corresponding to the facial image data; wherein, dynamic detection includes: blinking dynamic detection and lip micro-movement detection; calculating the liveness score for the corresponding time interval of the facial dynamic frames, and determining the face authenticity based on the liveness score to obtain the true state of the facial image data; wherein, the true state includes: real and fake; if the true state is real, matching the user data corresponding to the facial feature vector to determine the user's security identity data.

[0023] In one embodiment, using a camera and supplementary lighting installed in the control room or entrance, a person stands approximately 0.5 meters in front of the device, and a 2-second video stream of facial recognition (a group of 30 or 60 frames) is continuously captured. For this group of images, facial appearance, eye opening and closing, lip opening and closing are recorded, and it is determined whether the person blinks or makes mouth movements.

[0024] Then, based on the temporal features of facial dynamic frames, the non-periodic detection of blink intervals is performed to calculate the coefficient of variation (CV) of the blink interval sequence as standard deviation / mean. A score of 1 is obtained if CV > 0.3 (non-periodicity is a live feature), and a score of 0 is obtained if CV ≤ 0.3.

[0025] The test detects changes in lip opening and closing. The variance of the lip opening and closing sequence is calculated. A variance > 0.02 earns 1 point, and a variance ≤ 0.02 earns 0 points.

[0026] Dynamic range detection of continuous frame feature vector distance involves calculating the fluctuation range (max-min) of the Euclidean distance sequence of facial feature vectors in each frame. A score >0.1 earns 1 point, and ≤0.1 earns 0 points. These three scores are accumulated; a score ≥2 indicates a genuine status, and a score <2 indicates spoofing. When a genuine status is identified, the person's ID is determined based on facial feature comparison. The person's role level and host passwords are retrieved from the local database, and the corresponding security identity data package (personnel ID, role level, host password mapping table, etc.) is assembled to obtain the user's security identity data. When a spoofing status is identified, the identification request is rejected, an anomaly log is recorded, and an alarm notification is pushed to the security management center. The same person is prohibited from re-initiating identification within 30 seconds.

[0027] Then, the collected blink intervals are used as the test objects. If two or more blink actions are detected, the interval between each blink is not fixed and varies greatly. Similarly, if it is a looping video, the blink interval is relatively fixed, and it is necessary to detect whether there is periodic repetition.

[0028] Finally, the liveness score for the corresponding time interval of the facial dynamic frames is calculated. The similarity score can be calculated by assigning scores; for example, 1 point is awarded for the absence of periodic repetition, 1 point for blinking, and 1 point for mouth movement. This determines the user's facial recognition data to be 3 points, confirming the authenticity of the facial image data. Based on facial feature comparison, the user's ID is identified. The user's role level and host passwords are retrieved from the local database. The corresponding identity data packet (user ID, role level, host passwords, etc.) is assembled to obtain the user's security identity data, eliminating the need to obtain a key to determine the specific identity.

[0029] Step 102: Based on the preset trigger rule base, perform verification condition matching on the security identity data to obtain the security operation queue.

[0030] For example, this application performs verification condition matching on the security identity data based on the identified person's identity and the current state of the security system, and automatically determines the next security operation to be performed, realizing the pre-setting of security operations, enabling users to initiate the corresponding security operations after facial recognition.

[0031] Specifically, based on a preset trigger rule base, the security identity data is verified by matching the verification conditions to obtain a security operation queue. This includes: parsing the user permissions corresponding to the security identity data and reading the current access status of the security host; based on the user permissions and the current access status, matching the corresponding rule parameters in the trigger rule base to determine the operation entries of the security host; wherein, the operation entries include: trigger conditions, host operation template, and operation priority; and sorting the operation entries in execution order to obtain the security operation queue.

[0032] In one embodiment, the real-time status of each alarm host includes arming, disarming, and alarm status; the user's security identity data includes personnel number, role level, passwords for each host, etc.; and the security system maintenance typically pre-configures an operation rule library based on security management needs.

[0033] First, personnel IDs and role levels include administrators, duty officers, and temporary visitors. In the pre-defined trigger rule base, each rule is a statement corresponding to a condition and an action, such as: "If the role is administrator, disarm all hosts" or "If the role is duty officer, disarm only the lobby host," which can be represented using any existing encoding format. The format of each rule in the trigger rule base is {rule number, condition list, action list, priority, mutual exclusion flag}. Each condition in the condition list is a triple (field name, comparison operator, comparison value), and available fields include role level, current time, and current host status. When multiple rules are matched simultaneously, they are executed in descending order of priority; mutually exclusive rules only execute the one with the highest priority.

[0034] Read the role level and current status of each host from the security identity data, and traverse the rule base from high to low priority. If the role is an administrator, generate disarm all hosts; if the role is a regular duty officer and the current time is working time, generate disarm the lobby hosts.

[0035] Similarly, if any host is in an alarm state and its role is administrator, an operation sequence for de-alarming and disarming will be generated.

[0036] Then, all operations generated by matching rules are sorted according to rule priority and operation dependencies, following the order of disarming first and then arming other areas. This sorting is based on security considerations: any personnel entering a armed area will trigger an alarm, and the target area must be disarmed before safe entry is permitted. Only then can the areas still requiring protection be rearmed. The operation interval between disarming and arming is typically less than 2 seconds. During this window, the area is directly monitored by on-site personnel instead of by the system.

[0037] Finally, the security operation queue corresponding to the above sorting is output. If no rule matches, an empty queue is output. By automatically converting personnel identities into a series of specific security tasks, the user is spared the steps of clicking the screen or pressing the keyboard one by one.

[0038] Step 103: Construct parameterized execution instructions corresponding to each operation item in the security operation queue through parameter template parsing.

[0039] For example, this application constructs parameterized execution instructions corresponding to each operation item in the security operation queue through parameter template parsing, and constructs standardized instructions independent of various types of hosts, thereby improving the applicability of security control.

[0040] Then, the personnel number, operation type and timestamp in the security identity data are concatenated, and the signature value is calculated using the HMAC-SHA256 algorithm and the device-level key (pre-distributed by the security management center). The signature value is then appended as a hexadecimal string to the session number field of the instruction.

[0041] In one embodiment, firstly, the first operation to be executed in the security operation queue is selected, and the corresponding password is searched in the password mapping table of the identity data packet according to the hostname. If the password is not found, the operation is terminated and an error is recorded.

[0042] It is important to note that the cryptographic signature only verifies the legitimacy of the instruction source (to prevent forgery). The password field still needs to be filled into the key sequence in plaintext to simulate the keyboard input behavior of a legitimate user. This is an inherent requirement of the security host keyboard protocol.

[0043] Finally, the obtained characters are assembled into the corresponding parameterized execution command: {Target host: X1 host, Operation type: disarm, Parameters: Partition number=1, Password=1234, Timestamp: 20xx-0x-2x 1x:3x:2x.123, Session number: a3f2c1}, which can be represented by other types of characters.

[0044] It is important to note that cryptographic signatures are used to hash and sign personnel numbers, roles, operation types, and timestamps to prevent forgery.

[0045] Step 104: Obtain the real-time arming status of the security host, and determine the security command data by dynamically filling the key symbol sequence based on the parameterized execution command and the real-time arming status.

[0046] For example, this application converts the unified parameterized execution instructions into a key symbol sequence and binary data frame that the host keyboard bus can recognize, based on the actual state of the target host and historical successful experience, and provides an initial transmission timing sequence, thereby realizing the dynamic filling of the key symbol sequence.

[0047] Key symbols refer to the logical key identifiers (number keys 0-9, function keys ARM / DISARM / CLEAR, etc.) defined in the security host keyboard protocol. These are mapped to binary codes recognizable by the keyboard bus through a key symbol-binary frame lookup table in the protocol plugin. "Dynamic sequence filling" refers to replacing the parameter values ​​(password numbers, partition numbers, etc.) in the parameterized execution instructions bit by bit with the placeholder positions in the key template to generate a complete key symbol sequence.

[0048] Specifically, based on parameterized execution commands and real-time arming status, security command data is determined through dynamic filling of key symbol sequences. This includes: acquiring security key templates and setting the corresponding signal transmission intervals for the security key templates; where the security key templates include: regular disarming templates, alarm-extinguishing disarming templates, and zone disarming templates; performing virtual keyboard key sequence matching on the real-time arming status to determine the security key template corresponding to the real-time arming status; where the real-time arming status includes: alarm status, multi-zone status, and default status; and supplementing the parameterized execution commands into the placeholders corresponding to the security key templates to obtain the security command data.

[0049] Furthermore, virtual keyboard key sequence matching is performed on the real-time arming status to determine the security key template corresponding to the real-time arming status. Specifically, this includes: when the real-time arming status is in an alarm state, matching an alarm deactivation / disarming template to obtain a first security key template; when the real-time arming status is in a multi-zone state, matching a zone disarming template to obtain a second security key template; when the real-time arming status is in a default state, matching a regular disarming template to obtain a third security key template; obtaining the template layout corresponding to the successfully executed templates in the historical execution templates, and based on the first, second, or third security key template, using the corresponding template layout as the key template corresponding to the real-time arming status to determine the security key template.

[0050] In one embodiment, firstly, the protocol plugin corresponding to the host is loaded according to the target host name in the parameterized execution instruction. The corresponding plugin pre-stores multiple key sequence templates available for the host in different states; the normal disarming template is a password plus a disarming key; the disarming template in the alarm state is an alarm silence key plus a password plus a disarming key.

[0051] By actively sending status query commands or using the most recently cached host status (alarming or normal), combined with the operation parameters in the parameterized execution command: partition number 1, user password 4321, the most suitable template is dynamically selected.

[0052] If the host is currently in an alarm state and the command requests disarming, the template with a de-alarm button will be automatically selected; if a non-default partition is specified, the template with the partition prefix will be selected; if multiple conditions are met simultaneously, the templates will be merged. The selection process also considers the success rate of different templates under the same host and operation in the historical success experience database, prioritizing templates with higher success rates. After replacing the placeholders in the template with the actual parameters, the specific key symbol sequence is obtained.

[0053] Then, look up the key symbol and binary frame mapping table in the plugin, convert each symbol into a binary data frame that the host bus can recognize, change the alarm silence key to 0xC0, the number "4" to 0x34, and the disarm key to 0x3F.

[0054] Finally, the default inter-frame delay parameter of 80 milliseconds is read from the plugin, and an instruction execution plan data structure containing a binary frame list, an inter-frame delay list, an expected response code (0xE0 for a successful response) and a timeout is generated to obtain the security button template.

[0055] Step 105: Through command response detection, adjust the frame interval of the timing parameters corresponding to the security command data to obtain the security host control data.

[0056] For example, this application uses command response detection to adjust the frame interval of the timing parameters corresponding to the security command data to obtain the security host control data, thereby realizing heterogeneous security collaborative control with keyless face recognition.

[0057] Specifically, through command response detection, the frame interval of the timing parameters corresponding to the security command data is adjusted to obtain the security host control data. This includes: setting the initial frame interval corresponding to the security command data and sending the security command data to the security host at the initial frame interval; after sending, switching to receiving mode, receiving the response code returned by the security host by listening to the keyboard bus, demodulating the response code into the corresponding byte stream to determine the response type; if the response type is successful, recording the currently used frame interval parameter combination as a successful experience, increasing the selection weight of this frame interval parameter combination under the same host and operation type in the future (selection weight = number of successful times of this combination / total number of executions of this operation type on this host), to determine the first security host control data; if the response type is a password error, terminating the response of the security command data and generating an identity mismatch report to the host management end; if the response type is a timing error, adjusting the initial frame interval until the response type corresponding to the initial frame interval is successful, to determine the second security host control data; and obtaining the security host control data based on the first or second security host control data.

[0058] Furthermore, after adjusting the frame interval of the timing parameters corresponding to the security command data through command response detection to obtain the security host control data, the method also includes: extracting the trigger features of the security host control data and matching the security rules corresponding to the trigger features; wherein, the trigger features include: host identifier, operation type, execution status, and failure reason classification; according to the security rules, determining the operation delay of the host's operation queue through security action sorting, and updating the security host control data based on the operation delay; inserting the updated security host control data into the operation queue to determine the host security operation log.

[0059] In one embodiment, the historical time-series experience database is first queried to see if the host has a better frame interval record. If so, this value is used to override the default latency in the plan. The historical time-series experience database is a key-value store structure, where key = (host identifier, operation type, partition number) and value = {optimal frame interval, success count, total execution count, most recent success timestamp}. The default value of 80ms is returned when there are no records initially. Records older than 365 days are automatically discarded.

[0060] For timing errors, a bidirectional step-by-step search strategy is used to adjust the frame interval: Centered on the current interval T0, attempts are made sequentially at T0+10ms, T0-10ms, T0+5ms, T0-5ms, T0+2ms, and T0-2ms, for a maximum of 7 transmissions. If a successful response is received after an adjustment, the successful interval value is stored in the historical timing experience database, and the optimal interval is updated using a weighted average. If all 7 attempts fail, the host is deemed currently uncontrollable, a timing anomaly report is generated, and the operation entry is skipped.

[0061] After sending the last frame, wait 200 milliseconds for the host to open a receive window, listen for the acknowledgment data on the bus, and compare the actual received bytes with the planned expected acknowledgment code.

[0062] If the results are completely identical, the execution is considered successful; if a clear error code (such as 0xF0) is received, the password is considered incorrect, and the process is terminated immediately without retrying; if a timeout occurs or invalid data is received, the timing is considered incorrect.

[0063] For timing errors, if the same operation experiences two consecutive timing errors, the frame interval is successively increased or decreased by steps such as 5 milliseconds, 10 milliseconds, etc., based on the default value. The same command is resent each time until a successful response is received. The successful interval value is then stored in the historical timing experience database as the interval record for that host. If successful, the delay value used in this instance is also included in the weighted average of the experience database, allowing the optimal parameters to gradually converge.

[0064] Finally, the success / failure status, failure reason classification (incorrect password / timing error / timeout), the original response bytes returned by the host, and the actual delay parameters used are integrated and sent back to the corresponding protocol plugins and timing experience library for updating the historical success experience library and host status cache.

[0065] The above are embodiments of the method proposed in this application. Based on the same inventive concept, embodiments of this application also provide a heterogeneous security collaborative control device based on face recognition, the structure of which is as follows: Figure 2 As shown.

[0066] Figure 2 This is a schematic diagram of the internal structure of a heterogeneous security collaborative control device based on face recognition, provided as an embodiment of this application. Figure 2 As shown, the device includes: At least one processor 201; And a memory 202 that is communicatively connected to at least one processor; The memory 202 stores instructions that can be executed by at least one processor. The instructions are executed by at least one processor 201 to enable at least one processor 201 to implement a heterogeneous security collaborative control method based on face recognition.

[0067] Some embodiments of this application provide corresponding to Figure 1 A non-volatile computer storage medium for heterogeneous security collaborative control based on face recognition is disclosed, which stores computer-executable instructions that, when executed, can realize a heterogeneous security collaborative control method based on face recognition.

[0068] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for IoT devices and media are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0069] The systems, media, and methods provided in this application are one-to-one correspondences. Therefore, the systems and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the systems and media will not be repeated here.

[0070] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0071] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0072] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0073] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0074] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0075] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0076] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0077] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0078] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A heterogeneous security collaborative control method based on face recognition, characterized in that, The method includes: Acquire the user's facial image data and perform continuous frame liveness detection on the facial image data to determine the user's security identity data; Based on a preset trigger rule base, the security identity data is matched for verification conditions to obtain a security operation queue, which includes disarming operation entries and arming operation entries. By parsing parameter templates, parameterized execution instructions corresponding to each operation item in the security operation queue are constructed. The real-time arming status of the security host is obtained, and the security command data is determined by dynamically filling the key symbol sequence according to the parameterized execution command and the real-time arming status. By detecting the command response, the frame interval of the timing parameters corresponding to the security command data is adjusted to obtain the security host control data.

2. The heterogeneous security collaborative control method based on face recognition according to claim 1, characterized in that, Perform continuous frame liveness detection on the facial image data to determine the user's security identity data, specifically including: Locate the facial region in the face image data and extract the facial feature vector corresponding to each frame in the facial region; Dynamic detection is performed on the facial feature vector to determine the facial dynamic frame corresponding to the facial image data; wherein, the dynamic detection includes: blinking dynamic detection and lip micro-movement detection; Liveness scoring is performed based on the temporal features of the facial dynamic frames to obtain the true state of the face image data; wherein, the true state includes: real and fake; the liveness score is determined by accumulating the scores of at least one of the following detections: non-periodic detection of blink interval, detection of lip opening and closing changes, and dynamic range detection of the distance between facial feature vectors in consecutive frames; when the accumulated score exceeds a preset liveness threshold, it is determined to be real, and when it is lower than the threshold, it is determined to be fake; If the real state is real, match the user data corresponding to the facial feature vector to determine the user's security identity data; If the actual state is a disguise, reject the current identification request, record the anomaly log, and prohibit the same person from re-initiating identification within a preset cooldown period.

3. The heterogeneous security collaborative control method based on face recognition according to claim 1, characterized in that, Based on a preset trigger rule base, the security identity data is matched for verification conditions to obtain a security operation queue, specifically including: Parse the user permissions corresponding to the security identity data and read the current access status of the security host; Based on the user permissions and the current access status, the corresponding rule parameters in the trigger rule base are matched to determine the operation items of the security host; wherein, the operation items include: trigger conditions, host operation template, and operation priority; The operation entries are sorted in execution order to obtain the security operation queue.

4. The heterogeneous security collaborative control method based on face recognition according to claim 1, characterized in that, Through parameter template parsing, parameterized execution instructions corresponding to each operation item in the security operation queue are constructed, specifically including: The parameters of each operation item in the security operation queue are parsed to determine the placeholder corresponding to the operation item. Save the fixed parameters in the placeholders and perform string replacement on the non-fixed parameters in the placeholders to obtain the host execution parameters; The security identity data is encrypted and signed, and the encrypted and signed security identity data is integrated with the host execution parameters to determine the parameterized execution instruction; wherein, the encryption and signing adopts the HMAC-SHA256 algorithm.

5. The heterogeneous security collaborative control method based on face recognition according to claim 1, characterized in that, Based on the parameterized execution instructions and the real-time arming status, security instruction data is determined through dynamic filling of key symbol sequences, specifically including: Obtain security button templates and set the corresponding signal transmission intervals for the security button templates; wherein, the security button templates include: a regular disarming template, an alarm-extinguishing disarming template, and a zone disarming template; The real-time arming status is matched with a virtual keyboard key sequence to determine the security key template corresponding to the real-time arming status; wherein, the real-time arming status includes: alarm status, multi-zone status, and default status; The parameterized execution instructions are added to the placeholders corresponding to the security button template to obtain the security instruction data.

6. The heterogeneous security collaborative control method based on face recognition according to claim 5, characterized in that, The real-time arming status is matched with a virtual keyboard key sequence to determine the security key template corresponding to the real-time arming status, specifically including: When the real-time arming state is in an alarm state, the alarm disarming template is matched to obtain the first security button template; When the real-time arming state is a multi-zone state, the zone disarming template is matched to obtain the second security button template; Obtain the historical successful key template sequence that is the same as the current host identifier and operation type from the historical execution experience library, and use the key arrangement order and inter-frame delay parameters in the historical successful key template sequence as the arrangement and timing reference of the security key template matching the current real-time arming state to determine the security key template.

7. The heterogeneous security collaborative control method based on face recognition according to claim 1, characterized in that, By detecting command responses, the frame interval of the timing parameters corresponding to the security command data is adjusted to obtain the security host control data, specifically including: Set the initial frame interval corresponding to the security command data, and send the security command data to the security host via the keyboard bus at the initial frame interval; After sending, switch to receiving mode, receive the response code returned by the security host by listening to the keyboard bus, and demodulate the response code into the corresponding byte stream to determine the response type; If the response type is successful, the currently used frame interval parameter combination is recorded as a successful experience, and the selection weight of this frame interval parameter combination is increased in subsequent selections under the same host and operation type, so as to determine the first security host control data. In the case of a timing error in the response type, a bidirectional step search strategy is used to adjust the initial frame interval; If the adjusted response type is successful, the success interval value is stored in the historical time series experience database and the optimal interval is updated in a weighted average manner to determine the control data of the second security host. If the response type fails after several adjustments, the host is determined to be currently uncontrollable, a timing anomaly report is generated, and the operation entry is skipped. The security host control data is obtained based on the first security host control data or the second security host control data.

8. The heterogeneous security collaborative control method based on face recognition according to claim 1, characterized in that, After adjusting the frame interval of the timing parameters corresponding to the security command data through command response detection to obtain the security host control data, the method further includes: Extract the trigger features from the security host control data and match the corresponding security rules; wherein, the trigger features include: host identifier, operation type, execution status, and failure reason classification; According to the security rules, the operation delay of the host's operation queue is determined by sorting the security actions, and the security host control data is updated based on the operation delay; The updated security host control data is inserted into the operation queue to determine the host security operation log.

9. A heterogeneous security collaborative control device based on face recognition, characterized in that, The device includes: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, which are executed by the at least one processor to enable the at least one processor to implement a heterogeneous security collaborative control method based on face recognition as described in any one of claims 1-8.

10. A non-volatile computer storage medium for heterogeneous security collaborative control based on face recognition, storing computer-executable instructions, characterized in that, When the computer-executable instructions are executed, they can realize the heterogeneous security collaborative control method based on face recognition as described in any one of claims 1-8.

Images