Civil air defense informatization management system

Abstract

The invention discloses a civil air defense informatization management system, and the system comprises a personnel management module which is used for obtaining the face image information and identity information of a management personnel of a civil air defense project, and binding the face image information of the management personnel with the corresponding identity information; the illegal intrusion monitoring module is used for acquiring monitoring video data of a key area in the civil air defense project, analyzing personnel and behaviors in the monitoring video data to obtain an identity recognition result and a personnel behavior classification result, judging whether the personnel are illegal intrusion personnel or not based on the identity recognition result and the personnel behavior classification result, and if yes, sending the judgment result to the server; if the person is judged to be an illegal invading person, early warning is carried out; the fire monitoring module is used for acquiring a thermal image inside the civil air defense project, analyzing the thermal image, and if a fire is detected, performing early warning and planning an optimal evacuation path; not only can the management personnel and external personnel be identified, but also whether the behavior of the management personnel is consistent with the level can be identified, so that the internal safety of the civil air defense project is further ensured.

Description

Technical Field

[0001] This invention relates to the field of civil defense technology, and in particular to a civil defense information management system. Background Technology

[0002] Civil defense projects refer to underground protective structures built independently to ensure the shelter of personnel and supplies, command and control of civil air defense, and medical rescue during wartime, as well as basements built in conjunction with above-ground buildings that can be used for air defense during wartime. In disasters or emergencies, civil defense projects often serve as temporary shelters for people; therefore, they are stocked with basic living supplies to ensure that those seeking refuge have sufficient resources to meet their basic living needs.

[0003] Civil defense projects may store supplies and important confidential information. Therefore, it is crucial to verify the identities of personnel entering the internal areas of civil defense projects to prevent unauthorized intrusion. However, traditional civil defense management systems typically only use facial recognition at the entrances and exits of civil defense projects to verify the identities of personnel, but lack the ability to identify the behavior of management personnel entering the internal areas and cannot detect abnormal behavior of management personnel. Summary of the Invention

[0004] In view of the above-mentioned prior art, the present invention provides a civil defense information management system, which mainly solves the technical problems existing in the background art.

[0005] To achieve the above objectives, the technical solution of this invention is implemented as follows: A civil defense information management system, comprising: The personnel management module is used to obtain the facial image information and identity information of the management personnel of civil defense projects, and bind the facial image information of the management personnel with the corresponding identity information; The illegal intrusion monitoring module is used to acquire monitoring video data of key areas inside civil defense projects, analyze the personnel and behaviors in the monitoring video data to obtain identity recognition results and personnel behavior classification results, and determine whether they are illegal intruders based on the identity recognition results and personnel behavior classification results. If they are determined to be illegal intruders, an early warning is issued. The fire monitoring module is used to acquire thermal images of the interior of civil defense projects, analyze the thermal images, issue an early warning if a fire is detected, and plan the optimal evacuation route.

[0006] Optionally, the step of obtaining the facial image information and identity information of the personnel in charge of civil defense projects, and binding the facial image information of the personnel with the corresponding identity information, includes: Acquire facial image information of management personnel in preset states, including face occlusion, face not occlusion, wearing a safety helmet, and not wearing a safety helmet; Obtain the identity information of the management personnel, including their name, gender, and job title; The facial image information and identity information of managers in different preset states are bound together.

[0007] Optionally, the personnel in the surveillance video data are analyzed to obtain identification results, including: A facial recognition model is constructed based on the YOLO-V8 network. The facial image and identity information are input into the facial recognition model for training to obtain a trained facial recognition model. Real-time monitoring video data of key areas inside civil defense projects is acquired, and the real-time monitoring video data is segmented to obtain a frame sequence image dataset. The frame sequence image dataset is input into the trained facial recognition model to perform facial recognition and output the person's identity recognition result.

[0008] Optionally, the step of analyzing the behavior in the surveillance video data to obtain the personnel behavior classification result includes: The acquired surveillance video data is processed by setting the start frame position, end frame position, and time interval. The video surveillance data is then segmented to obtain a frame sequence image dataset, and the timestamp of each frame sequence image is recorded. According to the timestamp order, for each frame of the frame sequence image dataset, the position of the human body in each frame is detected based on the human bounding box of the YOLO-V8 algorithm; Extract the human motion trajectory from the human target bounding box; The human movement trajectory is used to extract the human's movement speed, direction change frequency, and dwell time in prohibited areas in consecutive frames, and these are then spliced ​​together to obtain comprehensive behavioral features. The comprehensive behavioral features are then labeled to generate tags. The comprehensive behavioral features and the generated labels constitute a behavioral dataset. The behavioral dataset is then input into a support vector machine for training to classify human behaviors and output the human behavior classification results.

[0009] Optionally, the step of determining whether someone is an unauthorized intruder based on the identity recognition result and the personnel behavior classification result, and issuing an early warning if the person is determined to be an unauthorized intruder, includes: If the identity verification result does not match the identity information of the corresponding person, it is determined to be an unauthorized intruder, and an alert is triggered; If the identity recognition result matches the identity information of the corresponding person, and the person's behavior classification result is abnormal, then the person is determined to be an unauthorized intruder, and an alert is triggered. If the identity recognition result matches the corresponding person's identity information and the person's behavior classification result is normal, then the person is determined to be a normal person and no warning is issued.

[0010] Optionally, the step of acquiring a thermal image of the interior of the civil defense project, analyzing the thermal image, and issuing an early warning if a fire is detected includes: Multiple temperature-sensitive optical fibers are installed inside the civil defense project area. These optical fibers are equidistant from each other and are labeled. A real-time thermal image of the internal monitoring area of ​​the civil defense facility is obtained, the center temperature of the real-time thermal image is extracted, and a temperature threshold and a time threshold are set. If the center temperature exceeds the temperature threshold and the duration exceeds the time threshold, a fire is determined to exist, and an early warning is issued.

[0011] Optionally, the planning of the optimal evacuation route includes: Obtain environmental data and an internal area map of the civil defense project. The environmental data includes ambient temperature data, carbon monoxide concentration data, and visibility data. Divide the interior of the civil defense project into multiple nodes based on the internal area map. Based on the environmental data, calculate the environmental risk level of each node; Calculate the path cost based on the environmental risk level and node spacing; Based on Dijkstra's algorithm, starting from the starting node, the path cost is iteratively calculated, the priority queue is updated, and the shortest path from the starting node to the target node is found.

[0012] Optionally, the expression for the risk level is:

[0013] in, For risk level, Relative temperature This refers to the relative carbon monoxide concentration. For relative visibility, As a weight for relative temperature, As a weight relative to carbon monoxide concentration, This is the weight for relative visibility.

[0014] Optionally, the expression for the path cost is:

[0015] in, For the node i To the nodei+1 distance, For nodes i The level of risk, n The number of nodes on the path. To adjust the coefficient for the impact of risk level on path cost.

[0016] The beneficial effects of this invention are as follows: The civil defense information management system provided by this invention, through a personnel management module and an illegal intrusion monitoring module, can combine identity recognition and behavior recognition to identify external personnel and management personnel to determine whether they are management personnel inside the civil defense project, thereby determining whether they are illegal intruders. Furthermore, it can also identify the behavior of management personnel to determine whether their behavior pattern is normal or abnormal. If the behavior pattern is abnormal, the management personnel are also determined to be illegal intruders. It can not only identify management personnel and external personnel, but also identify whether the behavior of management personnel matches their rank, further ensuring the security inside the civil defense project. Through the fire monitoring module, it can detect fires in the internal areas of the civil defense project, thereby planning the optimal evacuation route based on the fire point, facilitating faster and safer evacuation. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the structure of a civil defense information management system provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. In the following description, the expression "some embodiments" refers to a subset of all possible embodiments; however, it should be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments and can be combined with each other without conflict.

[0019] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0020] It should be understood that the present invention can be embodied in various forms and should not be construed as being limited to the embodiments set forth herein. Rather, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Furthermore, the terminology used herein is intended only to describe particular embodiments and is not intended to limit the invention. When used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the terms “compose” and / or “comprising,” when used in this specification, identify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups. When used herein, the term “and / or” includes any and all combinations of the associated listed items.

[0021] It should also be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "inner," "outer," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0022] To fully understand this invention, a detailed structure will be presented in the following description to illustrate the technical solution proposed by this invention. Optional embodiments of the invention are described in detail below; however, in addition to these detailed descriptions, the invention may have other embodiments.

[0023] Example Please refer to the attached document. Figure 1 This application provides a civil defense information management system, comprising: The personnel management module is used to obtain the facial image information and identity information of the management personnel of civil defense projects, and bind the facial image information of the management personnel with the corresponding identity information; The illegal intrusion monitoring module is used to acquire monitoring video data of key areas inside civil defense projects, analyze the personnel and behaviors in the monitoring video data to obtain identity recognition results and personnel behavior classification results, and determine whether they are illegal intruders based on the identity recognition results and personnel behavior classification results. If they are determined to be illegal intruders, an early warning is issued. The fire monitoring module is used to acquire thermal images of the interior of civil defense projects, analyze the thermal images, issue an early warning if a fire is detected, and plan the optimal evacuation route.

[0024] Specifically, the personnel management module first acquires and binds facial image information and identity information of the administrator. This facilitates subsequent identification using the bound facial image information and identity information in the illegal intrusion monitoring module. After identity verification confirms the administrator as an administrator, the module then identifies the administrator's behavior to verify whether the behavior conforms to the normal behavior patterns of administrators. If abnormal behavior is detected, it may indicate that the administrator is behaving in a manner inconsistent with their rank, such as lingering in a prohibited area for an extended period, thus identifying them as an illegal intruder. Therefore, this invention not only verifies external personnel to determine if they are administrators (identifying them as illegal intruders if they are not), but also tracks the behavior of administrators to determine if they exhibit abnormal behavior. If abnormal behavior is detected, they are also identified as illegal intruders. By combining identity recognition and behavior recognition, the system can more comprehensively manage and monitor personnel entering civil defense projects, ensuring the safety of the civil defense project interior. The system also includes a fire monitoring module. By detecting fires, if a fire is detected, the system can plan the optimal evacuation route based on the fire location for rapid evacuation.

[0025] As an optional implementation, the step of obtaining the facial image information and identity information of the personnel in charge of civil defense projects, and binding the facial image information of the personnel with the corresponding identity information, includes: Acquire facial image information of management personnel in preset states, including face occlusion, face not occlusion, wearing a safety helmet, and not wearing a safety helmet; Obtain the identity information of the management personnel, including their name, gender, and job title; The facial image information and identity information of managers in different preset states are bound together.

[0026] Specifically, different states are preset, and facial image information of each manager in all different states is collected. For example, facial images in the face occlusion state can include wearing glasses, not wearing glasses, wearing a mask, and not wearing a mask; facial images under different lighting conditions can also be collected. By collecting facial images in different preset states, more facial feature information of the managers can be obtained.

[0027] As an optional implementation, analyzing the personnel in the surveillance video data to obtain identification results includes: A facial recognition model is constructed based on the YOLO-V8 network. The facial image and identity information are input into the facial recognition model for training to obtain a trained facial recognition model. Real-time monitoring video data of key areas inside civil defense projects is acquired, and the real-time monitoring video data is segmented to obtain a frame sequence image dataset. The frame sequence image dataset is input into the trained facial recognition model to perform facial recognition and output the person's identity recognition result.

[0028] Specifically, the acquired real-time monitoring video data is segmented to obtain a frame sequence image dataset. Images without visible face areas are deleted, and facial features are extracted using the face target detection box based on the YOLO-V8 model. The extracted facial features are then compared and identified. If the corresponding management personnel are identified, the identity information of the corresponding management personnel will be output, thereby confirming that the person is a manager of the civil defense project.

[0029] As an optional implementation, the step of analyzing the behavior in the surveillance video data to obtain the personnel behavior classification result includes: The acquired surveillance video data is processed by setting the start frame position, end frame position, and time interval. The video surveillance data is then segmented to obtain a frame sequence image dataset, and the timestamp of each frame sequence image is recorded. According to the timestamp order, for each frame of the frame sequence image dataset, the position of the human body in each frame is detected based on the human bounding box of the YOLO-V8 algorithm; Extract the human motion trajectory from the human target bounding box; The human movement trajectory is used to extract the human's movement speed, direction change frequency, and dwell time in prohibited areas in consecutive frames, and these are then spliced ​​together to obtain comprehensive behavioral features. The comprehensive behavioral features are then labeled to generate tags. The comprehensive behavioral features and the generated labels constitute a behavioral dataset. The behavioral dataset is then input into a support vector machine for training to classify human behaviors and output the human behavior classification results.

[0030] Specifically, the sequential frame images are input into the model for processing. The model outputs a human target bounding box containing the human body. Based on the human target bounding box, the center point coordinates and the size of the bounding box are calculated to extract the human body's position information. By comparing the changes in the human body's position in consecutive frames, the movement speed of the human body is calculated. The movement trajectory of the human body in the human target bounding box in consecutive frames is extracted to determine the human body's movement direction and calculate the frequency of change of the human body's direction. Based on a preset prohibited area, it is determined whether the human body stays in the prohibited area and the dwell time is calculated. The calculated speed, direction change frequency, and dwell time are concatenated to obtain a comprehensive behavioral feature vector. The comprehensive behavioral feature vector is labeled to generate labels, which can be normal or abnormal. The comprehensive behavioral feature vector and the corresponding labels are then combined to form a behavioral dataset. A support vector machine is used to train the behavioral dataset to obtain a behavioral classification model. Real-time monitoring video data is input into the trained behavioral classification model to classify human behavior and output the human behavior classification results.

[0031] As an optional implementation, the step of determining whether someone is an unauthorized intruder based on the identity recognition result and the personnel behavior classification result, and issuing an early warning if the intruder is determined to be an unauthorized intruder, includes: If the identity verification result does not match the identity information of the corresponding person, it is determined to be an unauthorized intruder, and an alert is triggered; If the identity recognition result matches the identity information of the corresponding person, and the person's behavior classification result is abnormal, then the person is determined to be an unauthorized intruder, and an alert is triggered. If the identity recognition result matches the corresponding person's identity information and the person's behavior classification result is normal, then the person is determined to be a normal person and no warning is issued.

[0032] Specifically, after identifying a person and obtaining the identification result, a preliminary determination is made based on the identification result to determine whether the person is a manager. Then, the behavior of the person is identified. If no corresponding identification result is output, it means that the person is not a manager inside the corresponding civil defense project, but an external person. In this case, it is no longer necessary to judge their behavior; they are identified as an unauthorized intruder and an alert is issued. If the identification result outputs the corresponding identity information, it means that the person is a manager inside the civil defense project. In this case, it is still necessary to identify their behavior. Since managers of different management levels have different permissions and the areas they are allowed to enter also differ, after identifying them as managers, their behavior is then identified to determine whether the manager has any abnormal behavior. If abnormal behavior is found, they are also identified as an unauthorized intruder. By combining the identification results and the behavior identification results, a more comprehensive judgment of unauthorized intruders can be made. At the same time, if the person is identified as an unauthorized intruder and is a manager, a faster response can be made based on their identity information, reducing the potential losses caused by unauthorized intrusion.

[0033] As an optional implementation, the step of acquiring a thermal image of the interior of the civil defense project, analyzing the thermal image, and issuing an early warning if a fire is detected includes: Multiple temperature-sensitive optical fibers are installed inside the civil defense project area. These optical fibers are equidistant from each other and are labeled. A real-time thermal image of the internal monitoring area of ​​the civil defense facility is obtained, the center temperature of the real-time thermal image is extracted, and a temperature threshold and a time threshold are set. If the center temperature exceeds the temperature threshold and the duration exceeds the time threshold, a fire is determined to exist, and an early warning is issued.

[0034] Specifically, a fire is identified when both the center temperature and the duration exceed a set threshold. Thermal imaging analysis can identify areas of abnormal temperature, and combined with temperature fiber optic monitoring data, the location of the temperature fiber optic cable can be pinpointed. Based on the number of abnormal temperature fiber optic cables, the fire's coverage area can be determined, the fire's location can be identified, and an early warning can be issued. By combining temperature and time thresholds for judgment, false alarms due to short-term temperature fluctuations can be avoided.

[0035] As an optional implementation, the planning of the optimal evacuation route includes: Obtain environmental data and an internal area map of the civil defense project. The environmental data includes ambient temperature data, carbon monoxide concentration data, and visibility data. Based on the internal area map of the civil defense project, divide the interior of the civil defense project into multiple edges and multiple nodes. Based on the environmental data, calculate the environmental risk level of each edge; Calculate the path cost based on the environmental risk level and node spacing; Based on Dijkstra's algorithm, starting from the starting node, the path cost is iteratively calculated, the priority queue is updated, and the shortest path from the starting node to the target node is found.

[0036] Specifically, a structural diagram of the internal area of ​​the civil defense project is obtained, and an undirected graph is constructed based on the diagram. Channels are considered edges between nodes, and edges represent paths from one node to another. Intersections between channels are also considered nodes, and exits in the undirected graph are defined as nodes as well. Corresponding environmental data is collected using appropriate sensors, and the environmental risk level of each edge is calculated based on this data. For example, edges with excessively high temperatures, high carbon monoxide concentrations, or low visibility have higher risk levels. The path cost from one node to another is calculated based on the environmental risk level of each edge and the physical distance between nodes. Then, Dijkstra's algorithm is used to calculate the path cost. The algorithm finds the shortest path from the starting node to the target node as follows: Set the path cost of the starting node to 0 and the path costs of all other nodes to infinity. Create a priority queue to store nodes to be processed. Select the node with the lowest path cost from the priority queue and update the path costs of its adjacent nodes. If the path cost from the current node to its adjacent node is lower, update the path cost of the adjacent node and add it to the priority queue. After processing each node, update the priority queue to ensure that the nodes in the queue are sorted in ascending order of path cost. When the target node is processed or the priority queue is empty, find the shortest path from the starting node to the target node.

[0037] As an optional implementation, the expression for the risk level is:

[0038] in, For risk level, Relative temperature This refers to the relative carbon monoxide concentration. For relative visibility, As a weight for relative temperature, As a weight relative to carbon monoxide concentration, This is the weight for relative visibility.

[0039] As an optional implementation, the expression for the path cost is:

[0040] in, For the node i To the node i+1 distance, For nodes i The level of risk, n The number of nodes on the path. To adjust the coefficient for the impact of risk level on path cost.

[0041] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A civil air defense information management system, characterized in that, The method comprises the steps of: a personnel management module, configured to acquire face image information and identity information of a management personnel of the civil air defense project, and bind the face image information of the management personnel with corresponding identity information; an illegal intrusion monitoring module, configured to acquire monitoring video data of a key area inside the civil air defense project, analyze personnel and behavior in the monitoring video data to obtain an identity recognition result and a personnel behavior classification result, judge whether the personnel is illegal intrusion personnel based on the identity recognition result and the personnel behavior classification result, and if the personnel is judged to be illegal intrusion personnel, perform early warning; a fire monitoring module, configured to acquire a thermal imaging map inside the civil air defense project, analyze the thermal imaging map, and if a fire is detected, perform early warning and plan an optimal evacuation path.

2. The civil defense information management system of claim 1, wherein, The method comprises the steps of: acquiring face image information of the management personnel in a preset state, wherein the preset state comprises face occlusion, unoccluded face, wearing a safety helmet, and not wearing a safety helmet; acquiring identity information of the management personnel, wherein the identity information comprises name, gender, and job level; correspondingly binding the face image information and the identity information of the management personnel in different preset states.

3. The civil defense information management system of claim 1, wherein, The method comprises the steps of: constructing a face recognition model based on a YOLO-V8 network, inputting the face image and the identity information into the face recognition model for training, and obtaining a trained face recognition model; acquiring real-time monitoring video data of a key area inside the civil air defense project, performing segmentation processing on the real-time monitoring video data to obtain a frame sequence image data set; inputting the frame sequence image data set into the trained face recognition model, performing face recognition, and outputting an identity recognition result of the personnel.

4. The civil defense information management system of claim 1, wherein, The method comprises the steps of: processing the acquired monitoring video data, setting a start frame position, an end frame position, and a time interval, performing segmentation processing on the video monitoring data to obtain a frame sequence image data set, and recording a timestamp of each frame sequence image; based on a human body target box of a YOLO-V8 algorithm, detecting a position of a human body in each frame image in the frame sequence image data set according to a timestamp order; extracting a human body motion trajectory in the human body target box; extracting a moving speed, a direction change frequency, and a stay time in a prohibited area of the human body in consecutive frames from the human body motion trajectory, and splicing to obtain comprehensive behavior features, labeling the comprehensive behavior features, and generating labels; constructing a behavior data set from the comprehensive behavior features and the generated labels, inputting the behavior data set into a support vector machine for training, classifying personnel behavior, and outputting a personnel behavior classification result.

5. The civil defense information management system according to any one of claims 3 or 4, wherein, The method comprises the steps of: If the identity recognition result is that no corresponding personnel identity information is matched, it is determined that the intruder is illegal, and a warning is triggered; If the identity recognition result is that corresponding personnel identity information is matched, and the personnel behavior classification result is abnormal, it is determined that the intruder is illegal, and a warning is triggered; If the identity recognition result is that corresponding personnel identity information is matched, and the personnel behavior classification result is normal, it is determined that the intruder is normal, and no warning is performed.

6. The civil defense information management system of claim 1, wherein, The method comprises the following steps: A plurality of temperature optical fibers are arranged in the internal area of the civil air defense project, the plurality of temperature optical fibers are arranged at equal intervals, and each temperature optical fiber is labeled; A real-time thermal image of the internal monitoring area of the civil air defense project is obtained, the center temperature of the real-time thermal image is extracted, a temperature threshold and a time threshold are set, if the center temperature exceeds the temperature threshold and the duration exceeds the time threshold, it is determined that there is a fire, and a warning is performed.

7. The civil defense information management system of claim 6, wherein, The method for planning an optimal evacuation path comprises the following steps: Obtain environmental data information and an internal area map of the civil air defense project, the environmental data information comprises environmental temperature data, carbon monoxide concentration data, and visibility data; the internal area of the civil air defense project is divided into a plurality of nodes according to the internal area map of the civil air defense project; Based on the environmental data information, the environmental risk level of each node is calculated; According to the environmental risk level and the node distance, the path cost is calculated; Based on the Dijkstra algorithm, the path cost is iteratively calculated from the starting node, the priority queue is updated, and the shortest path from the starting node to the target node is found.

8. The civil defense information management system of claim 7, wherein, The expression of the risk level is: wherein, is a risk level, is a relative temperature, is a relative carbon monoxide concentration, is a relative visibility, is a weight for the relative temperature, is a weight for the relative carbon monoxide concentration, is a weight for the relative visibility.

9. The civil defense information management system of claim 8, wherein, The expression of the path cost is: wherein, is the distance from the node i to the node i+1 , is the risk level of the node i , n is the number of nodes on the path, is a coefficient that adjusts the impact of the risk level on the cost of the path.

Images