Intelligent monitoring system for secret-containing carrier and monitoring method thereof

Abstract

This invention provides an intelligent monitoring system and method for classified carriers, comprising: real-time perception and behavior monitoring of personnel or vehicles entering the area to obtain target object and behavior status information; when a target object is detected entering a classified area, scanning and verifying the electronic tag carried by the target object to generate a comprehensive verification result; determining the passage status of personnel or vehicles based on the comprehensive verification result, and uploading the passage status and corresponding verification result to a central server for recording and management. By jointly analyzing the verification results of the target object's behavior status information and tag identity information, and determining the corresponding feature acquisition strategy based on the credibility result and behavior status information, dynamic monitoring and hierarchical verification of personnel or vehicles are achieved. This improves upon the lack of comprehensive analysis of the correlation between the behavior status and identity information of entrants in existing technologies, which leads to a mismatch between verification results and actual risks.

Description

Technical Field

[0001] This application relates to the technical field of intelligent monitoring, and in particular to an intelligent monitoring system and monitoring method for classified carriers. Background Technology

[0002] With the continuous advancement of informatization and digitalization, various classified carriers are increasingly widely used in government agencies, research institutions, military industrial units, and critical infrastructure. Classified carriers typically contain important data or sensitive information, and their storage, transmission, and use require extremely high security. Unauthorized access, exchange, or removal of classified carriers during management or use can easily lead to information leaks and trigger serious security risks.

[0003] Intelligent monitoring of classified areas involves installing access control devices at the entrances to verify the identity of personnel or vehicles entering; using electronic tags, card readers, or biometric devices to confirm the identity of those entering; and simultaneously, coordinating with video surveillance systems to record and track activities within the classified areas, thereby achieving control over entry and exit behavior in classified areas.

[0004] However, while electronic tag identification, access control, and video surveillance can achieve basic management and recording of people entering and leaving confidential areas, in practical applications, when tags are misused or borrowed, or when the behavior of the person entering is abnormal but the identity verification result still shows that it is legitimate, existing technologies often make a passage judgment based on only a single identity verification result, lacking a comprehensive analysis of the relationship between the behavior status of the person entering and the identity information, thus leading to the problem of mismatch between the verification result and the actual risk. Summary of the Invention

[0005] To address the lack of comprehensive analysis of the relationship between the behavior status and identity information of entrants in existing technologies, which leads to a mismatch between verification results and actual risks, this application provides an intelligent monitoring system and method for classified carriers. By introducing a comprehensive judgment mechanism based on identity verification results and behavior status information, it achieves refined monitoring and dynamic control of personnel or vehicles entering classified areas, thereby improving the overall security level of classified areas.

[0006] In view of this, a first aspect of the present invention provides an intelligent monitoring method for classified carriers, applied to classified areas of classified carriers, comprising: real-time perception and behavior monitoring of personnel or vehicles entering the classified area using monitoring equipment deployed in the classified area to obtain target object and behavior status information; when the target object is detected entering the classified area, scanning the electronic tag carried by the target object to obtain tag identity information, and verifying the tag identity information to generate a credibility result; determining a corresponding feature acquisition strategy based on the credibility result and the behavior status information to perform local feature acquisition of the personnel or vehicle; verifying the consistency of the association between the tag identity information and the local features to generate a comprehensive verification result; determining the passage status of the personnel or vehicle based on the comprehensive verification result, and uploading the passage status and the corresponding verification result to a central server for recording and management.

[0007] Optionally, the step of real-time perception and behavior monitoring of personnel or vehicles entering the classified area using monitoring equipment deployed in the classified area includes: detecting the thermal radiation characteristics of the entering target using an infrared thermal imaging sensor array installed in the entrance passage of the classified area to identify the presence of personnel or vehicles; acquiring the movement speed, movement trajectory, and volume parameters of the entering target using a millimeter-wave radar sensor installed in the entrance passage of the classified area; performing fusion analysis on the thermal radiation characteristics, the movement speed, the movement trajectory, and the volume parameters to determine whether the target object is a personnel or a vehicle; and using a high-definition camera deployed on the side wall of the passage to collect dynamic image sequences of personnel or vehicles, and performing behavior analysis on the dynamic image sequences to extract behavioral state information.

[0008] Optionally, the step of scanning the electronic tag carried by the target object to obtain tag identity information and verifying the tag identity information to generate a credibility result when the target object is detected entering the confidential area includes: when the target object is detected to enter a preset sensing range, activating an RFID reader deployed at the entrance to send a query signal to the electronic tag and receiving the tag identity information fed back by the electronic tag; wherein, the tag identity information includes a unique identifier, permission level, validity period, and encrypted signature; and comparing the unique identifier with a list of authorized identifiers pre-stored in a local security database. The system performs a comparison to confirm the legitimacy of the tag; it uses a preset asymmetric encryption algorithm to decrypt and verify the encrypted signature to determine the authenticity of the tag; it checks whether the validity period is within the current time range to determine the timeliness of the tag; it compares the permission level with the minimum access permission required by the classified area to determine the permission matching degree; it calculates a comprehensive credibility score based on the verification results of the tag legitimacy, tag authenticity, tag timeliness, and permission matching degree; when the comprehensive credibility score is higher than a preset threshold, a high credibility result is generated; when the comprehensive credibility score is lower than the preset threshold, a low credibility result is generated.

[0009] Optionally, the step of determining the corresponding feature acquisition strategy based on the credibility result and the behavioral state information to perform local feature acquisition on the person or vehicle includes: when the credibility result is a high credibility result and no abnormal action features appear in the behavioral state information, activating a normal resolution camera to acquire a single frame image of the person's facial area or the vehicle's license plate area; when the credibility result is a low credibility result or abnormal action features appear in the behavioral state information, activating a high-definition multi-angle camera array to acquire multiple frames of continuous images of the person's facial features, gait features, and iris features, or the vehicle's license plate features, body color features, and vehicle model outline features; when the behavioral state information indicates that the person or vehicle is carrying items, activating an X-ray scanning device to perform perspective imaging on the carried items to obtain feature images of the items' internal structure; preprocessing the acquired feature images, and extracting local feature vectors from the preprocessed feature images and encoding and storing them.

[0010] Optionally, the step of verifying the consistency between the tag identity information and the local features to generate a comprehensive verification result includes: retrieving a pre-registered associated biometric template or vehicle feature template from the central server based on the unique identifier in the tag identity information; calculating the matching degree value between the local feature vector and the associated biometric template or vehicle feature template; determining that the identity features are consistent when the matching degree value is greater than a preset consistency threshold; determining that the identity features are inconsistent when the matching degree value is less than or equal to the consistency threshold; detecting the time difference between the reading timestamp of the electronic tag and the collection timestamp of the local feature; generating a comprehensive verification result that passes verification when the identity features are consistent and the time difference is within a preset time range; and generating a comprehensive verification result that fails verification when the identity features are inconsistent or the time difference is not within the preset time range.

[0011] Optionally, the step of determining the passage status of the personnel or vehicle based on the comprehensive verification result and uploading the passage status and the corresponding verification result to the central server for recording and management includes: when the comprehensive verification result is a successful verification result, sending an opening command to the entrance control device to allow the personnel or vehicle to pass, and setting the passage status to a permitted passage status; when the comprehensive verification result is a failed verification result, sending a rejection command to the entrance control device to prevent the personnel or vehicle from passing, setting the passage status to a rejected passage status, and triggering an audible and visual alarm device for on-site warning; encrypting and encapsulating the tag identity information, the local features, the behavioral status information, the credibility result, the comprehensive verification result, the passage status, and the current timestamp to obtain a verification record data packet, and uploading the verification record data packet to the central server for recording and management through a secure communication channel.

[0012] Optionally, after receiving the verification record data packet, the central server decrypts and verifies its integrity, stores the verified records in a distributed database and establishes an index; for the records in the denied access state, it triggers an abnormal event analysis process and pushes an early warning notification to the security management personnel.

[0013] A second aspect of this invention provides an intelligent monitoring system for classified carriers, applied to classified areas of the classified carriers, comprising: an area monitoring module, used to perform real-time perception and behavior monitoring of personnel or vehicles entering the classified area through monitoring equipment deployed in the classified area, to obtain target object and behavior status information; an identity verification module, used to scan the electronic tag carried by the target object when the target object is detected entering the classified area, to obtain tag identity information, and to verify the tag identity information to generate a credibility result; a feature acquisition module, used to determine a corresponding feature acquisition strategy based on the credibility result and the behavior status information, to perform local feature acquisition of the personnel or vehicles; a feature verification module, used to verify the consistency of the association between the tag identity information and the local features, to generate a comprehensive verification result; and a passage control module, used to determine the passage status of the personnel or vehicles according to the comprehensive verification result, and to upload the passage status and the corresponding verification result to a central server for recording and management.

[0014] The technical solution of the present invention has the following advantages: by jointly analyzing the verification results of the target object's behavioral state information and tag identity information, and determining the corresponding feature collection strategy based on the credibility results and behavioral state information, dynamic monitoring and hierarchical verification of personnel or vehicles are realized. Furthermore, by verifying the consistency of the association between tag identity information and local features, the accuracy and reliability of the comprehensive verification results are improved. This addresses the problem in the prior art of lacking a comprehensive analysis of the relationship between the behavioral state and identity information of the entering object, which leads to a mismatch between the verification results and the actual risks. Attached Figure Description

[0015] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating an intelligent monitoring method for classified carriers provided in an embodiment of the present invention; Figure 2 This is a schematic block diagram of the structure of an intelligent monitoring system for classified carriers provided in an embodiment of the present invention.

[0017] Figure label: 10. An intelligent monitoring system for classified carriers; 11. Area monitoring module; 12. Identity verification module; 13. Feature acquisition module; 14. Feature verification module; 15. Access control module. Detailed Implementation

[0018] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] In the description of this invention, it should be noted that the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0020] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0021] like Figure 1 As shown in the example of this application, an intelligent monitoring method for classified carriers is provided. Applied to the classified areas of these carriers, it enables multi-dimensional perception, identity verification, and behavioral analysis of personnel or vehicles entering the classified areas, thereby forming a comprehensive monitoring mechanism based on target object and behavioral status information. This intelligent monitoring method for classified carriers specifically includes steps S1 to S5: Step S1: Real-time perception and behavior monitoring of personnel or vehicles entering the area are carried out using monitoring equipment deployed in classified areas to obtain information on the target objects and their behavior status.

[0022] Data is collected from the targets entering the classified area by monitoring devices arranged in the entrance passage, side walls, and top of the passage. The collected data is processed to form target object and behavior status information. The target object is used to characterize the type of object entering the classified area, and the behavior status information is used to characterize the movement and behavior characteristics of the target object in the classified area.

[0023] For example, in the entrance passage of a classified area, when the monitoring equipment collects continuously changing spatial data and thermal radiation data, the target object entering the classified area is determined to be a person or a vehicle based on the combined characteristics of the spatial data and thermal radiation data. At the same time, behavioral status information is extracted based on the changes in the continuously collected data.

[0024] In another example, step S1 may also be preferably performed as follows: The presence of personnel or vehicles is identified by detecting the thermal radiation characteristics of targets entering the area through an array of infrared thermal imaging sensors installed at the entrance passage of the classified area.

[0025] The infrared thermal imaging sensor array samples the thermal radiation of the entrance channel area. The sampling period is set to 50 milliseconds to 200 milliseconds. A two-dimensional thermal map is constructed from the sampled thermal radiation data. The area, contour shape and temperature gradient parameters of the thermal radiation distribution area are extracted. The presence status of the entering target is determined based on the spatial size and temperature gradient characteristics of the thermal radiation area.

[0026] When the area parameter of the thermal radiation zone is between 0.3 square meters and 1.5 square meters and the temperature gradient shows a continuous change, it is determined that there is a human target; when the area parameter of the thermal radiation zone is greater than 3 square meters and the temperature gradient distribution shows a multi-peak characteristic, it is determined that there is a vehicle target.

[0027] Millimeter-wave radar sensors installed at the entrance to the classified area acquire the speed, trajectory, and volume parameters of the incoming target.

[0028] Millimeter-wave radar sensors continuously detect targets within the entrance channel, acquiring their distance, velocity, and angle information. The sensors calculate the target's velocity, trajectory, and volume parameters using multiple frames of radar echo data. The velocity is calculated from the target's displacement change per unit time, the trajectory is composed of the target's continuous position coordinates, and the volume parameters are calculated from the target's radar cross-section.

[0029] When the millimeter-wave radar sensor detects that the target's speed is between 0.2 m / s and 2 m / s and the movement trajectory shows continuous path characteristics, the target is determined to be a person; when the millimeter-wave radar sensor detects that the target's speed is greater than 2 m / s and the volume parameter is greater than a set threshold, the target is determined to be a vehicle.

[0030] The thermal radiation characteristics, movement speed, movement trajectory and volume parameters are fused and analyzed to determine whether the target object is a person or a vehicle.

[0031] Feature vectors are constructed from thermal radiation characteristics, movement speed, movement trajectory, and volume parameters to form target feature vectors. The target feature vectors are then classified using a support vector machine (SVM) algorithm. The pre-trained classification model is used to classify the target feature vectors and output the target object category results.

[0032] When the target feature vector is classified as a person by the SVM classification model, the target object is determined to be a person; when the target feature vector is classified as a vehicle by the SVM classification model, the target object is determined to be a vehicle.

[0033] High-definition cameras deployed on the side walls of the passageway are used to collect dynamic image sequences of people or vehicles. Behavioral analysis is performed on the dynamic image sequences to extract behavioral status information, including movement direction, dwell time, abnormal action characteristics, and the status of carried items.

[0034] The high-definition camera captures dynamic image sequences at a frame rate of 25 to 60 frames per second. Target detection and tracking are performed on the dynamic image sequences. The YOLO target detection algorithm is used for target detection, and the KCF target tracking algorithm is used for target tracking. The movement direction and dwell time are calculated based on the target's position changes in consecutive frames. Abnormal action features are identified based on changes in target posture and trajectory. The status of carried items is identified based on the target's contour extension features and the contours of additional objects.

[0035] When the target's position change direction in consecutive frames is consistent with the entrance channel direction, the movement direction is determined as the entry direction; when the target's position change in consecutive frames is less than a set threshold and the duration is greater than 3 seconds, the dwell time is determined as abnormal dwell; when the target's attitude change amplitude exceeds a set threshold, the abnormal action feature is determined as abnormal behavior; when an additional contour area appears outside the target contour, the carrying item status is determined as carrying items.

[0036] Step S2: When a target object is detected entering a classified area, the electronic tag carried by the target object is scanned to obtain the tag identity information, and the tag identity information is verified to generate a credibility result.

[0037] The electronic tags carried by the target object are scanned by the radio frequency identification reading and writing device set up at the entrance channel of the classified area to obtain the tag identity information, the tag identity information is parsed and verified, and a credibility result is generated based on the verification result.

[0038] When the RFID reader receives the RFID response signal from the electronic tag, it decodes the RFID response signal to obtain the tag's identity information and verifies the tag's identity information according to preset verification rules.

[0039] In another example, step S2 may also be preferably performed as follows: When a target object is detected to enter the preset sensing range, the RFID reader deployed at the entrance is activated to send a query signal to the electronic tag and receive the tag identity information fed back by the electronic tag. The tag identity information includes a unique identifier, access level, validity period and encrypted signature.

[0040] When a target object enters the preset sensing range, the RFID reader sends a query signal. After receiving the query signal, the electronic tag returns a data frame containing a unique identifier, access level, validity period, and encrypted signature. The tag's identity information is obtained by parsing the data frame.

[0041] For example, the unique identifier uses a 128-bit encoding format, the permission level uses a multi-level permission encoding method, the validity period is represented by a timestamp format, and the encrypted signature uses a digital signature generated based on the RSA algorithm.

[0042] The unique identifier is compared with a list of authorized identifiers pre-stored in the local security database to confirm the label's legitimacy.

[0043] The unique identifier is matched against the identifiers in the list of authorized identifiers. The matching method is an exact string matching method. When the unique identifier is exactly the same as any identifier in the list of authorized identifiers, the label is deemed to be valid.

[0044] The label is deemed valid when the unique identifier matches one of the identifiers in the list of authorized identifiers; the label is deemed invalid when the unique identifier does not appear in the list of authorized identifiers.

[0045] The encrypted signature is decrypted and verified using a preset asymmetric encryption algorithm to determine the authenticity of the tag.

[0046] The encrypted signature is decrypted using the RSA asymmetric encryption algorithm. The public key is used to decrypt the encrypted signature to obtain the signature digest information. The signature digest information is then compared with the digest information generated from the tag identity information. If the two match, the tag is deemed to be authentic.

[0047] The tag is considered authentic when the decrypted signature digest matches the tag identity digest; otherwise, the tag is considered unauthentic.

[0048] Check whether the expiration date is within the current time frame to determine the label's timeliness.

[0049] The validity period is compared with the current time. If the current time is within the time range specified by the validity period, the label is deemed valid.

[0050] The label is deemed valid when the current time is greater than the start time of the validity period but less than the end time; the label is deemed invalid when the current time exceeds the time range of the validity period.

[0051] The permission level is compared with the minimum access permission required for classified areas to determine the degree of permission matching.

[0052] The permission level is compared with the minimum access permission required for the classified area. When the permission level is greater than or equal to the minimum access permission, the permission matching degree is determined to be satisfactory.

[0053] When the permission level is level 3 and the minimum access permission is level 2, the permission matching degree is determined to be satisfied; when the permission level is level 1 and the minimum access permission is level 2, the permission matching degree is determined to be unsatisfactory.

[0054] A comprehensive credibility score is calculated based on the verification results of label legality, label authenticity, label timeliness, and permission matching.

[0055] The validity, authenticity, timeliness, and permission matching of the labels are quantitatively assigned values. The quantitative assignment adopts a binary scoring method or a multi-level scoring method, and the comprehensive credibility score is obtained by weighted summation of each score.

[0056] For example, the label validity is assigned a value of 1 or 0, the label authenticity is assigned a value of 1 or 0, the label timeliness is assigned a value of 1 or 0, and the permission matching degree is assigned a value of 1 or 0. The values ​​of each item are weighted and summed according to the weights of 0.3, 0.3, 0.2, and 0.2 to obtain the comprehensive credibility score.

[0057] When the overall credibility score is higher than the preset threshold, a high credibility result is generated.

[0058] The overall credibility score is compared with a preset threshold. When the overall credibility score is greater than the preset threshold, the credibility result is determined to be a high credibility result.

[0059] When the overall credibility score is 0.85 and the preset threshold is 0.7, the credibility result is determined to be a high credibility result.

[0060] When the overall credibility score is lower than the preset threshold, a low credibility result is generated.

[0061] The overall credibility score is compared with a preset threshold. When the overall credibility score is less than or equal to the preset threshold, the credibility result is determined to be a low credibility result.

[0062] When the overall credibility score is 0.5 and the preset threshold is 0.7, the credibility result is determined to be a low credibility result.

[0063] Step S3: Determine the corresponding feature acquisition strategy based on the credibility results and behavioral status information to collect local features of personnel or vehicles.

[0064] By constructing a joint decision rule for credibility results and behavioral state information, the credibility results and behavioral state information are mapped to corresponding feature acquisition strategies. The feature acquisition strategies include low-precision acquisition strategies, high-precision acquisition strategies, and extended acquisition strategies. The feature acquisition strategy corresponding to the target object is determined by a preset strategy matching table.

[0065] For example, the credibility results are divided into high credibility results and low credibility results. Abnormal action features and the status of carried items in the behavior status information are used as policy triggering conditions to form a correspondence table of "credibility result - behavior status information - feature collection strategy". When the credibility result is a high credibility result and no abnormal action features appear in the behavior status information, a low-precision collection strategy is matched. When the credibility result is a low credibility result or abnormal action features appear in the behavior status information, a high-precision collection strategy is matched. When the behavior status information contains the status of carried items, an extended collection strategy is matched.

[0066] In another example, step S3 may also be preferably performed as follows: When the credibility result is high and no abnormal action features appear in the behavior status information, a low-precision acquisition strategy is adopted, and a normal resolution camera is activated to acquire single-frame images of the person's face area or the vehicle's license plate area.

[0067] By controlling the trigger frequency, exposure parameters, and frame rate of a standard resolution camera, a single-frame image is captured when the target object enters the preset capture area of ​​a confidential area. The capture area is then dynamically cropped to limit the image range of the person's face or the vehicle's license plate area.

[0068] For example, the resolution of a standard resolution camera can be set to 1280×720 pixels, the trigger condition can be set to the target object entering the preset acquisition area and the target object's moving speed being less than the preset speed threshold, and the acquisition area can be limited to the area corresponding to the face detection box or license plate detection box in the target object image.

[0069] When the credibility result is low or abnormal action features appear in the behavioral status information, a high-precision acquisition strategy is adopted, and a high-definition multi-angle camera array is activated to acquire multiple frames of continuous images of the facial features, gait features, and iris features of the person or the license plate features, body color features, and vehicle outline features of the vehicle.

[0070] By controlling the synchronous triggering mechanism of the high-definition multi-angle camera array, multiple high-definition cameras can capture the target object from multiple perspectives within the same time window, and the number of consecutive capture frames and the capture time length can be set to form a multi-frame continuous image sequence of the target object.

[0071] For example, the resolution of the high-definition multi-angle camera array is set to 3840×2160 pixels, the number of continuous acquisition frames is set to no less than 30 frames, the acquisition time is set to no less than 1 second, and the acquisition angle is set to frontal view, side view and oblique side view, in order to obtain facial features, gait features and iris features of people or license plate features, body color features and vehicle outline features of vehicles.

[0072] When the behavior status information indicates that a person or vehicle is carrying items, an additional X-ray scanning device is activated to perform a transillumination imaging of the carried items in order to obtain characteristic images of the internal structure of the items.

[0073] By setting the scanning area of ​​the X-ray scanning device to correspond with the area of ​​the object being carried, and controlling the scanning energy, scanning angle, and scanning time of the X-ray scanning device, a perspective image of the object being carried can be formed.

[0074] For example, the scanning energy of the X-ray scanning device is set to 80keV to 160keV, the scanning area is limited to the spatial range corresponding to the items carried by the target object, and the scanning time is set to 0.5 seconds to 2 seconds to obtain characteristic images of the internal structure of the items carried.

[0075] The preprocessing of the acquired feature images includes denoising, enhancement, and normalization operations, and local feature vectors are extracted from the preprocessed feature images and encoded and stored.

[0076] The stability of the feature image is improved by performing Gaussian filtering for noise reduction, histogram equalization for enhancement, and pixel intensity normalization on the feature image. Local feature vectors are extracted from the preprocessed feature image based on the feature extraction algorithm and encoded into fixed-length feature descriptors.

[0077] For example, Gaussian filtering algorithm is used to denoise the feature image, histogram equalization algorithm is used to enhance the feature image, Min-Max normalization algorithm is used to normalize the feature image, local binary mode algorithm or scale-invariant feature transform algorithm is used to extract local feature vectors, the local feature vectors are encoded into 128-dimensional or 256-dimensional feature descriptors and stored in the local feature cache.

[0078] Step S4: Verify the consistency between the tag identity information and local features to generate a comprehensive verification result.

[0079] By constructing a mapping relationship between tag identity information and local features, the identity template corresponding to the tag identity information is matched and analyzed with the local features, and a comprehensive verification result is generated based on the matching degree value and time consistency condition.

[0080] For example, the unique identifier in the tag identity information is used as the index key, and the local feature vector is used as the matching object. The corresponding associated biometric template or vehicle feature template is located by the unique identifier, and the matching degree between the local feature vector and the associated biometric template or vehicle feature template is calculated.

[0081] In another example, step S4 may also be preferably performed as follows: The pre-registered associated biometric template or vehicle feature template is retrieved from the central server based on the unique identifier in the tag's identity information.

[0082] By using a unique identifier as a query condition, a template query request is sent to the central server, and the associated biometric template or vehicle feature template is returned by the central server.

[0083] For example, the unique identifier is encoded into a string and a query request is sent to the central server through a secure communication protocol. The central server then returns a template of facial features, gait features, and iris features for individuals, or a template of license plate features, vehicle body color features, and vehicle model outline features for vehicles.

[0084] The matching degree between the local feature vector and the feature vector of the associated biometric template or vehicle feature template is calculated using the Euclidean distance algorithm or the cosine similarity algorithm.

[0085] By mapping the local feature vector to the feature vector of the associated biometric template or vehicle feature template into the same feature space, and using the Euclidean distance algorithm or cosine similarity algorithm to calculate the distance or similarity value between the two.

[0086] For example, the Euclidean distance algorithm can be used to calculate the distance between the local feature vector and the feature vector of the associated biometric template or vehicle feature template, and the distance value can be converted into a matching degree value. Alternatively, the cosine similarity algorithm can be used to calculate the similarity between the local feature vector and the feature vector of the associated biometric template or vehicle feature template, and the similarity value can be used as the matching degree value.

[0087] When the matching score is greater than the preset consistency threshold, the identity features are considered to be consistent.

[0088] By comparing the matching score with a preset consistency threshold, the system outputs a result indicating that the identity features are consistent when the matching score is greater than the consistency threshold.

[0089] For example, if the consistency threshold is set to 0.85, a match score of 0.92 is considered to indicate consistent identity characteristics.

[0090] When the matching degree value is less than or equal to the consistency threshold, it is determined that the identity features are inconsistent.

[0091] By comparing the matching score with a consistency threshold, the system outputs a result indicating inconsistency in identity features when the matching score does not meet the condition of being greater than the consistency threshold.

[0092] For example, if the consistency threshold is set to 0.85, a match score of 0.63 is considered an inconsistency in identity features.

[0093] The time difference between the reading timestamp of the electronic tag and the local feature acquisition timestamp is detected.

[0094] By obtaining the reading timestamp and local feature acquisition timestamp of the electronic tag, and calculating the time difference between the two, the temporal correlation between the tag identity information and the local features can be determined.

[0095] For example, let the timestamp for reading the electronic tag be T1, and the timestamp for acquiring local features be T2. Calculate the time difference ΔT = |T2|. T1|.

[0096] When the identity features are consistent and the time difference is within the preset time range, a comprehensive verification result that passes the verification is generated.

[0097] By logically combining the identity feature consistency judgment result and the time difference judgment result, when the identity features are consistent and the time difference meets the preset time range condition, the comprehensive verification result that passes the verification is output.

[0098] For example, if the preset time range is set to 0 to 3 seconds, a comprehensive verification result that passes verification will be generated when the identity features are consistent and ΔT is 1.2 seconds.

[0099] When identity features are inconsistent or the time difference is outside the preset time range, a comprehensive verification result that fails verification is generated.

[0100] By logically combining the identity feature inconsistency judgment result with the time difference judgment result, when the identity feature is inconsistent or the time difference does not meet the preset time range condition, the comprehensive verification result of failing verification is output.

[0101] When the identity features are inconsistent or ΔT is 6.5 seconds, a comprehensive verification result that fails verification is generated.

[0102] Step S5: Determine the passage status of personnel or vehicles based on the comprehensive verification results, and upload the passage status and corresponding verification results to the central server for recording and management.

[0103] By constructing a mapping rule between comprehensive verification results and access status, the comprehensive verification results are mapped to corresponding access control instructions, and verification record data packets are generated synchronously, thereby realizing the synchronous execution of physical access control and information record management.

[0104] For example, the comprehensive verification results are divided into those that pass verification and those that fail verification, and a correspondence table of "comprehensive verification result - passage status - control command" is established. When the comprehensive verification result is a successful verification result, an enable command is generated, and when the comprehensive verification result is a failed verification result, a deny command is generated.

[0105] In another example, step S5 may also be preferably performed as follows: When the comprehensive verification result is a pass, an opening command is sent to the entrance control device to allow personnel or vehicles to pass, and the passage status is set to the allowed passage status.

[0106] The execution status of the entry control device is controlled by encoding the start command into a control command data frame and sending it to the entry control device through a wired or wireless communication interface.

[0107] For example, the opening instruction can be encoded into a control instruction data frame containing a target object identification field, an access permission field, and an instruction type field, and sent to the entry control device via the RS-485 communication protocol or the TCP / IP communication protocol, so that the entry control device can perform the opening action.

[0108] When the comprehensive verification result is a failure, a rejection command is sent to the entrance control device to prevent personnel or vehicles from passing, the passage status is set to a rejection status, and the audible and visual alarm device is triggered to issue an on-site warning.

[0109] By encoding the rejection instruction into a control instruction data frame and sending it to the entry control device through the communication interface, and simultaneously sending a trigger signal to the audible and visual alarm device, the audible and visual alarm device is put into an alarm state.

[0110] For example, the rejection instruction is encoded into a control instruction data frame containing a target object identification field, a rejection reason field, and an instruction type field, and sent to the entry control device via the RS-485 communication protocol or the TCP / IP communication protocol. The alarm mode of the audible and visual alarm device is set to a high-frequency flashing mode and a continuous sound mode.

[0111] The tag identity information, local features, behavioral status information, credibility results, comprehensive verification results, access status and current timestamp are encrypted and encapsulated to obtain a verification record data packet, which is then uploaded to the central server for recording and management through a secure communication channel.

[0112] By constructing a data structure for the verification record data packet, the tag identity information, local features, behavior status information, credibility result, comprehensive verification result, passage status, and current timestamp are encapsulated in a preset field order, and encrypted using both symmetric and asymmetric encryption algorithms to ensure the data security of the verification record data packet.

[0113] For example, the verification record data packet is defined as JSON or binary TLV format, the tag identity information is mapped to the Tag_ID field, the local features are mapped to the Feature_Vector field, the behavior status information is mapped to the Behavior_State field, the credibility result is mapped to the Credibility_Result field, the comprehensive verification result is mapped to the Verification_Result field, the passage status is mapped to the Pass_Status field, and the current timestamp is mapped to the Timestamp field. The verification record data packet is symmetrically encrypted using the AES algorithm, the AES key is encrypted using the RSA algorithm, and the verification record data packet is transmitted through the TLS communication protocol.

[0114] Meanwhile, after receiving the verification record data packet, the central server decrypts and verifies its integrity, stores the verified records in the distributed database, and creates an index.

[0115] The verification record data packets are decrypted and their integrity is verified using a message digest algorithm to confirm that they have not been tampered with. The verified records are then written to the distributed database.

[0116] For example, the RSA algorithm is used to decrypt the AES key, the AES algorithm is used to decrypt the verification record data packet, the SHA-256 algorithm is used to generate a digest value for the verification record data packet and compare it, the verified record is stored in the verification record table of the distributed database, and an index is created using the tag identity information, the current timestamp, and the passage status as index fields.

[0117] The record of the denied passage status triggers the abnormal event analysis process and sends an early warning notification to the security management personnel.

[0118] By constructing abnormal event judgment rules, the records of denied passage status are matched with preset abnormal event rules, and warning information is generated based on the matching results. The warning notification is then pushed to security management personnel through a preset notification channel.

[0119] For example, the records of denied passage status can be correlated with thresholds for the number of times abnormal action features occur, thresholds for the number of times low-confidence results occur, and time window parameters. When the number of times the same tag identity information is denied passage status within a preset time window exceeds a preset threshold, an abnormal event identifier is generated, and an early warning notification is pushed through SMS interface, instant messaging interface, or security management platform interface.

[0120] In this application example, monitoring equipment deployed in the classified area performs real-time perception and behavior monitoring of personnel or vehicles entering the area, obtaining target object and behavior status information. When the target object is detected entering the classified area, the electronic tag carried by the target object is scanned to obtain tag identity information, and the tag identity information is verified to generate a credibility result. Subsequently, based on the credibility result and the behavior status information, a corresponding feature collection strategy is determined to collect local features of the personnel or vehicle. After completing the local feature collection, the consistency between the tag identity information and the local features is verified to generate a comprehensive verification result. Finally, the passage status of the personnel or vehicle is determined according to the comprehensive verification result, and the passage status and the corresponding verification result are uploaded to the central server for recording and management, thereby realizing intelligent monitoring and security control of personnel or vehicles in the classified area of ​​the classified carrier.

[0121] By jointly analyzing the behavioral state information of the target object with the verification results of the tag identity information, and determining the corresponding feature collection strategy based on the credibility results and behavioral state information, dynamic monitoring and hierarchical verification of personnel or vehicles are realized. This allows the local feature collection process to be triggered differently according to different risk states, thereby avoiding the misjudgment problem caused by single identity verification. At the same time, by verifying the consistency of the association between tag identity information and local features, the accuracy and reliability of the comprehensive verification results are improved. Combined with the mechanism of determining the passage status and uploading to the central server for recording and management based on the comprehensive verification results, refined control and traceable management of access behavior in classified areas are realized. This improves the problem in existing technologies that lack comprehensive analysis of the relationship between the behavioral state and identity information of the entrant, which leads to the mismatch between verification results and actual risks. Thus, it effectively improves the security protection level and the intelligence level of monitoring in classified areas where classified carriers are located.

[0122] like Figure 2 As shown, in another example, this application also provides an intelligent monitoring system 10 for classified carriers, applied to classified areas of classified carriers. The system mainly includes the following modules: The area monitoring module 11 is used to perform real-time perception and behavior monitoring of personnel or vehicles entering the area through monitoring equipment deployed in the classified area, and to obtain information on the target object and its behavior status.

[0123] The identity verification module 12 is used to scan the electronic tag carried by the target object when the target object is detected to enter the confidential area, obtain the tag identity information, and verify the tag identity information to generate a credibility result.

[0124] The feature acquisition module 13 is used to determine the corresponding feature acquisition strategy based on the credibility result and behavior status information, so as to perform local feature acquisition of personnel or vehicles.

[0125] The feature verification module 14 is used to verify the consistency between the tag identity information and local features in order to generate a comprehensive verification result.

[0126] The access control module 15 is used to determine the access status of personnel or vehicles based on the comprehensive verification results, and upload the access status and corresponding verification results to the central server for recording and management.

[0127] In this application example, the area monitoring module 11 performs real-time perception and behavior monitoring of personnel or vehicles within the classified area, acquiring target object and behavior status information; the identity verification module 12 scans and verifies the electronic tags carried by the target objects, comprehensively evaluating the legality, authenticity, timeliness, and permission matching degree of the tag identity information, thereby generating a credibility result; the feature acquisition module 13 determines the corresponding feature acquisition strategy based on the credibility result and behavior status information, achieving hierarchical and differentiated acquisition of local features of personnel or vehicles; the feature verification module 14 verifies the consistency of the association between tag identity information and local features, achieving a comprehensive judgment on the consistency of target object identity features and time consistency, thereby generating a comprehensive verification result; the access control module 15 determines the access status of personnel or vehicles based on the comprehensive verification result, and uploads the access status and corresponding verification result to the central server for recording and management, realizing the coordinated linkage of identity verification, behavior monitoring, feature acquisition, association verification, and access control of personnel or vehicles within the classified area, thereby improving the intelligent monitoring accuracy, identity verification reliability, and security management traceability of classified carriers.

[0128] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and each module described above can be referred to the corresponding process in the aforementioned Embodiment 1, and will not be repeated here.

[0129] The above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. An intelligent monitoring method for classified carriers, applied to classified areas of classified carriers, characterized in that, include: By deploying monitoring equipment in the classified area, real-time perception and behavior monitoring of personnel or vehicles entering the area are conducted to obtain information on target objects and their behavioral status. When the target object is detected to enter the classified area, the electronic tag carried by the target object is scanned to obtain the tag identity information, and the tag identity information is verified to generate a credibility result. Based on the credibility result and the behavioral state information, a corresponding feature acquisition strategy is determined to perform local feature acquisition on the personnel or vehicles. The consistency between the tag identity information and the local features is verified to generate a comprehensive verification result; The passage status of the personnel or vehicles is determined based on the comprehensive verification results, and the passage status and corresponding verification results are uploaded to the central server for recording and management.

2. The intelligent monitoring method for classified carriers according to claim 1, characterized in that, The step of using monitoring equipment deployed in the classified area to perform real-time perception and behavior monitoring of personnel or vehicles entering the area includes: The presence of personnel or vehicles is identified by detecting the thermal radiation characteristics of targets entering the classified area through an array of infrared thermal imaging sensors installed at the entrance passage. The speed, trajectory, and volume parameters of the incoming target are obtained by millimeter-wave radar sensors installed at the entrance passage of the classified area. The thermal radiation characteristics, the movement speed, the movement trajectory, and the volume parameters are fused and analyzed to determine whether the target object is a person or a vehicle. High-definition cameras deployed on the side walls of the passageway are used to collect dynamic image sequences of people or vehicles, and behavioral analysis is performed on the dynamic image sequences to extract behavioral state information.

3. The intelligent monitoring method for classified carriers according to claim 1, characterized in that, The step of scanning the electronic tag carried by the target object when it is detected that the target object has entered the classified area, obtaining the tag identity information, and verifying the tag identity information to generate a credibility result includes: When the target object is detected to enter the preset sensing range, the radio frequency identification reader deployed at the entrance is activated to send a query signal to the electronic tag and receive the tag identity information fed back by the electronic tag; wherein, the tag identity information includes a unique identifier, access level, validity period and encrypted signature; The unique identifier is compared with a list of authorized identifiers pre-stored in the local security database to confirm the legality of the label; The encrypted signature is decrypted and verified using a preset asymmetric encryption algorithm to determine the authenticity of the tag; Check whether the validity period is within the current time range to determine the timeliness of the label; The permission level is compared with the minimum access permission required for the classified area to determine the degree of permission matching. A comprehensive credibility score is calculated based on the verification results of the tag's legality, tag's authenticity, tag's timeliness, and permission matching degree. When the overall credibility score is higher than a preset threshold, a high credibility result is generated; When the overall credibility score is lower than a preset threshold, a low credibility result is generated.

4. The intelligent monitoring method for classified carriers according to claim 3, characterized in that, The step of determining a corresponding feature acquisition strategy based on the credibility result and the behavioral state information to perform local feature acquisition on the personnel or vehicle includes: When the credibility result is a high credibility result and no abnormal action features appear in the behavior status information, the ordinary resolution camera is activated to collect single-frame images of the person's face area or the vehicle's license plate area. When the credibility result is a low credibility result or abnormal action features appear in the behavior status information, a high-definition multi-angle camera array is activated to acquire multiple frames of continuous images of the facial features, gait features and iris features of the person or the license plate features, body color features and vehicle outline features of the vehicle. When the behavior status information indicates that a person or vehicle is carrying items, the X-ray scanning equipment is activated to perform a transillumination imaging of the carried items in order to obtain a characteristic image of the internal structure of the items. The acquired feature images are preprocessed, and local feature vectors are extracted from the preprocessed feature images and encoded and stored.

5. The intelligent monitoring method for classified carriers according to claim 4, characterized in that, The step of verifying the consistency between the tag identity information and the local features to generate a comprehensive verification result includes: The pre-registered associated biometric template or vehicle feature template is retrieved from the central server based on the unique identifier in the tag identity information. Calculate the matching degree between the local feature vector and the associated biometric template or vehicle feature template; When the matching degree value is greater than the preset consistency threshold, it is determined that the identity features are consistent; When the matching degree value is less than or equal to the consistency threshold, it is determined that the identity features are inconsistent. Detect the time difference between the reading timestamp of the electronic tag and the local feature acquisition timestamp; When the identity features are consistent and the time difference is within the preset time range, a comprehensive verification result that passes the verification is generated; When identity features are inconsistent or the time difference is outside the preset time range, a comprehensive verification result that fails verification is generated.

6. The intelligent monitoring method for classified carriers according to claim 5, characterized in that, The steps of determining the passage status of the personnel or vehicles based on the comprehensive verification results, and uploading the passage status and corresponding verification results to the central server for recording and management, include: When the comprehensive verification result is a successful verification result, an opening command is sent to the entrance control device to allow the personnel or vehicles to pass, and the passage status is set to the passage permission status. When the comprehensive verification result is a failure, a rejection command is sent to the entrance control device to prevent the personnel or vehicles from passing, the passage status is set to a rejection status, and an audible and visual alarm device is triggered to issue an on-site warning. The tag identity information, the local features, the behavior status information, the credibility result, the comprehensive verification result, the passage status, and the current timestamp are encrypted and encapsulated to obtain a verification record data packet, which is then uploaded to the central server for recording and management through a secure communication channel.

7. The intelligent monitoring method for classified carriers according to claim 6, characterized in that, After receiving the verification record data packet, the central server decrypts and verifies its integrity, stores the verified records in the distributed database, and creates an index. The record of the denied passage status triggers an abnormal event analysis process and sends an early warning notification to the security management personnel.

8. An intelligent monitoring system for classified carriers, applied to classified areas of the classified carriers, characterized in that, include: The area monitoring module is used to perform real-time perception and behavior monitoring of personnel or vehicles entering the area through monitoring equipment deployed in the classified area, and to obtain target object and behavior status information; The identity verification module is used to scan the electronic tag carried by the target object when the target object is detected to enter the confidential area, obtain the tag identity information, and verify the tag identity information to generate a credibility result. The feature acquisition module is used to determine the corresponding feature acquisition strategy based on the credibility result and the behavior state information, so as to perform local feature acquisition on the personnel or vehicles. The feature verification module is used to verify the consistency between the tag identity information and the local features in order to generate a comprehensive verification result; The access control module is used to determine the access status of the personnel or vehicles based on the comprehensive verification results, and upload the access status and the corresponding verification results to the central server for recording and management.

Images