Multi-mode public safety behavior dynamic identification method, system, equipment and medium

Abstract

The invention relates to a multi-mode public safety behavior dynamic identification method, system and device and a medium. The method comprises the following steps: acquiring multi-modal monitoring data; based on the environment monitoring data, combining with the first public space model, and performing identification to obtain an ideal public safety behavior type of each second public space model; according to the personnel monitoring data and the environment monitoring data, identifying an actual public safety behavior type of the personnel corresponding to the personnel monitoring data; performing conformity judgment on the ideal public safety behavior type and the actual public safety behavior type to obtain a conformity judgment result; if the conformity judgment result is conformity, public safety behavior execution guidance information is generated; and if the conformity judgment result is departure, generating public safety behavior change guidance information. By adopting the method, the behavior guidance strategy can be adjusted according to the real-time environment change, and the response capability to emergencies is improved.

Description

Technical Field

[0001] This application belongs to the field of data identification and processing, and in particular relates to a method, system, device and medium for dynamic identification of multimodal public safety behaviors. Background Technology

[0002] In the field of public safety, timely perception and effective control of potential risks in public spaces are core requirements for ensuring the safety of public life and property. Early public safety management relied primarily on manual inspections, which struggled to achieve full-time, all-area coverage. Furthermore, risk assessment was highly dependent on human experience, making it prone to omissions or misjudgments. Traditional public safety behavior recognition methods also lacked real-time analysis of monitored information, often only allowing for post-incident tracing when sudden safety incidents occurred, hindering pre-incident warnings and in-process intervention. Moreover, traditional public safety behavior recognition models have significant limitations when dealing with densely populated and complex public spaces, making it difficult to address diverse and sudden safety risks.

[0003] With the rapid development of the Internet of Things and artificial intelligence, multi-source data fusion has become an important trend in public safety monitoring. By integrating various types of data and building a comprehensive perception network, combined with intelligent algorithms for analysis and processing, dynamic assessment of the safety status of public spaces can be achieved.

[0004] However, existing public safety behavior recognition technologies still face some challenges in practical applications. They lack dynamism and specificity, making it difficult to dynamically generate optimal public safety response behaviors based on the environmental characteristics of a specific space. In addition, the judgment criteria for public safety behaviors in existing technologies are relatively fixed, failing to deeply adapt to specific scenarios and failing to fully consider the impact of real-time environmental parameters, thus limiting the efficiency of emergency response. Summary of the Invention

[0005] Therefore, it is necessary to provide a multimodal public safety behavior dynamic recognition method, system, device, and medium that can dynamically identify the environment and make real-time behavioral decisions, addressing the aforementioned technical problems.

[0006] Firstly, this application provides a method for dynamic identification of multimodal public safety behaviors, including:

[0007] Acquire multimodal monitoring data, which includes personnel monitoring data and environmental monitoring data;

[0008] Based on environmental monitoring data and combined with the first public space model, the ideal public safety behavior types of each second public space model in the first public space model are identified.

[0009] Based on personnel monitoring data and environmental monitoring data, and combined with the second public space model, the actual public safety behavior types of the personnel corresponding to the personnel monitoring data are identified.

[0010] The conformity between ideal public safety behavior types and actual public safety behavior types is assessed to obtain the conformity assessment results.

[0011] If the compliance assessment result is compliant, public safety behavior execution guidance information is generated based on the ideal public safety behavior type;

[0012] If the compliance assessment result is a deviation, public safety behavior change guidance information is generated based on the actual public safety behavior type and the ideal public safety behavior type. The public safety behavior change guidance information is used to guide personnel to change from performing the actual public safety behavior type to performing the ideal public safety behavior type.

[0013] In one embodiment, the ideal public safety behavior type includes the ideal waiting-for-rescue behavior type and the ideal evacuation behavior type. Based on environmental monitoring data and combined with the first public space model, the ideal public safety behavior type of each second public space model in the first public space model is identified, including:

[0014] Based on environmental monitoring data, the passage hazard coefficients of each third public space model in the first public space model are identified. The third public space model is used to connect the first exit of the first public space model and the second exit of each second public space model.

[0015] The evacuation risk value of each second public space model is calculated based on the passage hazard coefficient of the third public space model corresponding to each second public space model.

[0016] If the evacuation risk value is less than the preset evacuation risk threshold, the ideal public safety behavior type will be set to the ideal evacuation behavior type.

[0017] If the evacuation risk value is greater than the preset evacuation risk threshold, the ideal public safety behavior type will be set to the ideal waiting-for-rescue behavior type.

[0018] In one embodiment, the environmental monitoring data includes video environmental monitoring data and smoke sensor monitoring data. Based on the environmental monitoring data, the access hazard coefficients of each third public space model in the first public space model are identified, including:

[0019] The obstacle distribution characteristics of the third public space model are identified based on video environmental monitoring data, and the obstacle risk coefficient is generated based on the obstacle distribution characteristics.

[0020] Based on the smoke sensor monitoring data, the smoke concentration characteristics and smoke type characteristics of the third public space model are identified, and the passage toxicity risk coefficient is generated based on the smoke concentration characteristics and smoke type characteristics.

[0021] The passage hazard coefficient is calculated by combining the passage obstacle hazard coefficient and the passage toxicity hazard coefficient.

[0022] In one embodiment, the expression for the evacuation risk value is:

[0023]

[0024] In the formula, R i Let i be the evacuation risk value for the second public space model with serial number i. D represents the total number of third common spaces corresponding to the second common space model with index i. i,j I represents the passage hazard factor of the third public space with index j corresponding to the second public space model with index i. i,j L is the shared influence coefficient generated for the third public space with index j corresponding to the second public space model with index i, based on the shared second public space model. i,j S represents the length of the evacuation route corresponding to the third public space with index j, which is the second public space model with index i. i,j The safety benchmark coefficient is the third public space with index j corresponding to the second public space model with index i. The safety benchmark coefficient is used to characterize the completeness of the safety facilities in the third public space.

[0025] In one embodiment, the actual public safety behavior type includes the actual rescue-awaiting behavior type corresponding to the ideal rescue-awaiting behavior type and the actual evacuation behavior type corresponding to the ideal evacuation behavior type. Based on personnel monitoring data and environmental monitoring data, combined with the second public space model, the actual public safety behavior type of the personnel corresponding to the personnel monitoring data is identified, including:

[0026] Based on environmental monitoring data and combined with the second public space model, identify the areas in the second public space model that require assistance.

[0027] Based on personnel monitoring data, combined with a personnel identification neural network model and a second public space model, the location information of personnel is identified.

[0028] If a person's location information belongs to an area awaiting assistance, set the actual public safety behavior type to the actual behavior awaiting assistance type;

[0029] If the location information of a person belongs to the fourth public space model, which is outside the area awaiting rescue in the second public space model, the actual public safety behavior type will be set to the actual evacuation behavior type.

[0030] In one embodiment, a conformity assessment is performed on the ideal public safety behavior type and the actual public safety behavior type to obtain a conformity assessment result, including:

[0031] If the actual public safety behavior type is the actual rescue-awaiting behavior type and the ideal public safety behavior type is the ideal rescue-awaiting behavior type, or the actual public safety behavior type is the actual evacuation behavior type and the ideal public safety behavior type is the ideal evacuation behavior type, the compliance judgment result will be set to compliance.

[0032] If the actual public safety behavior type is the actual evacuation behavior type and the ideal public safety behavior type is the ideal waiting-for-rescue behavior type, or if the actual public safety behavior type is the actual waiting-for-rescue behavior type and the ideal public safety behavior type is the ideal evacuation behavior type, the conformity judgment result will be set as deviation.

[0033] In one embodiment, if the compliance determination result is compliant, public safety behavior execution guidance information is generated based on the ideal public safety behavior type, including:

[0034] If the compliance judgment result is compliant, and the ideal public safety behavior type is the ideal evacuation behavior type, take the location information as the starting point and each second exit as the ending point, perform path planning, and generate the second public space evacuation distance for each second exit;

[0035] Based on the evacuation distance of the second public space of each second exit and the passage hazard coefficient of the third public space model corresponding to each second exit, the modified evacuation hazard coefficient of each second exit is generated.

[0036] The second exit with the lowest evacuation risk coefficient is set as the optimal evacuation exit, and public safety action execution guidance information is generated based on the optimal evacuation exit.

[0037] Secondly, this application also provides a multimodal public safety behavior dynamic recognition system, including:

[0038] The monitoring data acquisition module is used to acquire multimodal monitoring data, which includes personnel monitoring data and environmental monitoring data.

[0039] The ideal behavior analysis module is used to identify the ideal public safety behavior types of each second public space model in the first public space model based on environmental monitoring data and combined with the first public space model;

[0040] The actual behavior recognition module is used to identify the type of actual public safety behavior of personnel corresponding to the personnel monitoring data based on personnel monitoring data and environmental monitoring data, combined with the second public space model;

[0041] The compliance determination module is used to determine the compliance between ideal public safety behavior types and actual public safety behavior types, and obtain the compliance determination result.

[0042] The behavior execution guidance module is used to generate public safety behavior execution guidance information based on the ideal public safety behavior type if the compliance judgment result is compliant.

[0043] The behavior change guidance module is used to generate public safety behavior change guidance information based on the actual public safety behavior type and the ideal public safety behavior type if the compliance judgment result is a deviation. The public safety behavior change guidance information is used to guide personnel to change from performing the actual public safety behavior type to performing the ideal public safety behavior type.

[0044] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method as described in any of the first aspects of this application.

[0045] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method as described in any of the first aspects of this application.

[0046] The aforementioned multimodal public safety behavior dynamic identification methods, systems, equipment, and media, through dynamic environmental perception, can adjust behavioral guidance strategies in real time according to the dynamic evolution of emergencies, thereby improving the response capability to emergencies, enhancing the accuracy of emergency response, and strengthening the prediction and response capabilities to complex emergencies. By introducing a dynamic modeling mechanism to correlate environmental parameters with the first public space model in real time, it can ensure the real-time nature of the generated ideal behavior types, improve the matching degree between ideal behavior types and the physical environment, and enable ideal behavior types to accurately match the actual risk state of emergencies, providing a dynamic benchmark for subsequent behavioral guidance. By combining personnel monitoring data and environmental monitoring data to dynamically determine the actual public safety behavior types of personnel, it can achieve immediate detection of personnel behavior deviations, adapt to the uncertainty of personnel behavior in emergencies, and thus enable behavioral guidance to adapt to different scenario requirements in real time, improving the flexibility of public safety emergency response. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 A flowchart illustrating a multimodal public safety behavior dynamic recognition method provided in one embodiment of this application;

[0049] Figure 2 A flowchart illustrating another method for dynamic recognition of multimodal public safety behaviors provided in one embodiment of this application;

[0050] Figure 3 This is a schematic diagram of the structure of a multimodal public safety behavior dynamic recognition system provided in one embodiment of this application. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0052] In one embodiment, such as Figure 1 As shown, a multimodal public safety behavior dynamic recognition method is provided. This embodiment illustrates the application of this method to a public safety monitoring terminal. It is understood that this method can also be applied to a public safety monitoring server, and further to a system including both a public safety monitoring terminal and a public safety monitoring server, and is implemented through the interaction between the two. In this embodiment, the method includes the following steps:

[0053] Step S101: Obtain multimodal monitoring data.

[0054] Specifically, the public safety monitoring terminal can acquire multimodal monitoring data in the first public space corresponding to the first public space model collected by sensor devices connected to the public safety monitoring terminal via electrical and / or wireless connection. The multimodal monitoring data may include personnel monitoring data and environmental monitoring data.

[0055] Optionally, personnel monitoring data may include, but is not limited to, visible light video personnel monitoring data, infrared video personnel monitoring data, and audio personnel monitoring data, which can be used to analyze the location of personnel in the primary public space.

[0056] Optionally, environmental monitoring data may include, but is not limited to, visible light video environmental monitoring data, infrared video environmental monitoring data, audio environmental monitoring data, and smoke environmental monitoring data.

[0057] Step S102: Based on environmental monitoring data and combined with the first public space model, identify the ideal public safety behavior type of each second public space model in the first public space model.

[0058] Specifically, the public safety monitoring terminal can identify the ideal public safety behavior types of each second public space model within the first public space model by combining environmental monitoring data collected by sensor devices with the first public space model mounted on the terminal. The first public space model can be a global digital twin model covering the entire public space, representing the overall building structure, layout, equipment distribution, and cross-regional linkage logic. The second public space model can be a local digital twin unit contained within the first public space model, categorized by function or risk level, representing specific floors, zones, rooms, or passageways within the public space.

[0059] Optionally, the first and second public space models may include, but are not limited to, Building Information Modeling (BIM), Geographic Information System (GIS) models, and raster map models G(M,N) constructed based on BIM and GIS models. The raster weights of the raster map model G(M,N) can be used to characterize the accessibility of the spatial location corresponding to that raster, and the public safety monitoring terminal can update the raster weights of the raster map model G(M,N) in real time based on environmental monitoring data.

[0060] Optionally, the ideal public safety behavior type can be used to characterize the optimal public safety behavior type calculated by risk assessment algorithm and strategy optimization algorithm based on real-time environmental monitoring data in a specific second public space model. The ideal public safety behavior type can be used to guide personnel to take the safest public safety behavior in the event of a public safety emergency.

[0061] Furthermore, ideal public safety behavior types may include, but are not limited to, ideal waiting-for-rescue behavior types and ideal evacuation behavior types.

[0062] Step S103: Based on personnel monitoring data and environmental monitoring data, and combined with the second public space model, identify the actual public safety behavior type of the personnel corresponding to the personnel monitoring data.

[0063] Specifically, the public safety monitoring terminal can identify the location information P = (x, y, t, i) of personnel within the second public space model based on personnel monitoring data and the second public space model. The public safety monitoring terminal can also identify the rescue-needed area within the second public space model based on environmental monitoring data and the second public space model. The rescue-needed area is a relatively safe temporary refuge area within the second public space model. Here, i is the sequence number of the second public space model, and t is the time parameter. Let i be the row number of the raster map in the second common space model. Let i be the number of columns in the raster map of the second common space model.

[0064] Optionally, the public safety monitoring terminal can identify the hazard source based on environmental monitoring data, obtain the preset safe area in the second public space model, and identify the area to be rescued in the second public space model by combining the impact range of the hazard source and the safe area.

[0065] Step S104: Determine the degree of conformity between the ideal public safety behavior type and the actual public safety behavior type to obtain the conformity determination result.

[0066] Specifically, the public safety monitoring terminal can determine the degree of conformity between the ideal public safety behavior type in the second public space model and the actual public safety behavior type of the personnel in the second public space model, and obtain the degree of conformity determination result.

[0067] Optionally, when the ideal public safety behavior type of the second public space model and the actual public safety behavior type of the personnel in the second public space model are the same type of public safety behavior, the public safety monitoring terminal can set the compliance judgment result to "compliant"; when the ideal public safety behavior type of the second public space model and the actual public safety behavior type of the personnel in the second public space model are different types of public safety behavior, the public safety monitoring terminal can set the compliance judgment result to "deviation".

[0068] Step S105: If the compliance judgment result is compliant, generate public safety behavior execution guidance information based on the ideal public safety behavior type.

[0069] Specifically, when the compliance judgment result is compliant, the public safety monitoring terminal can generate public safety behavior execution guidance information based on the ideal public safety behavior type and the location information of personnel in the second public space model.

[0070] Step S106: If the compliance judgment result is a deviation, generate public safety behavior change guidance information based on the actual public safety behavior type and the ideal public safety behavior type. The public safety behavior change guidance information is used to guide personnel to change from performing the actual public safety behavior type to performing the ideal public safety behavior type.

[0071] In the aforementioned multimodal dynamic identification method for public safety behaviors, by collaboratively analyzing personnel monitoring data and environmental monitoring data, various dynamic information in the emergency environment can be captured in real time, providing a comprehensive and accurate basis for adjusting behavioral guidance strategies, improving the sensitivity to environmental changes, and thus enhancing the initial response speed to emergencies; by generating and adjusting ideal public safety behavior types in real time, behavioral guidance strategies can be adapted to real-time environmental changes, thereby effectively responding to emergencies.

[0072] In an optional embodiment of this application, the ideal public safety behavior type may include the ideal waiting-for-rescue behavior type and the ideal evacuation behavior type. Please refer to [reference needed]. Figure 1 and Figure 2 Step S102, based on environmental monitoring data and combined with the first public space model, identifies the ideal public safety behavior types for each second public space model within the first public space model, which may include:

[0073] Step S202: Based on environmental monitoring data, the access hazard coefficients of each third public space model in the first public space model are identified.

[0074] Optionally, the third public space model can be used to characterize the public space connecting the first exit of the first public space model and the second exits of each of the second public space models.

[0075] Optionally, the third public space model may include, but is not limited to, Building Information Modeling (BIM), Geographic Information System (GIS) models, and raster map models constructed based on BIM and GIS models. The access hazard coefficient may include local access hazard coefficients and global access hazard coefficients. The public safety monitoring terminal can identify the local access hazard coefficients of each grid cell in the raster map model of the third public space model within the first public space model based on environmental monitoring data, and identify the global access hazard coefficients of the third public space model based on the local access hazard coefficients.

[0076] Step S203: Calculate the evacuation risk value of each second public space model based on the passage hazard coefficient of the third public space model corresponding to each second public space model.

[0077] Optionally, the public safety monitoring terminal can calculate the evacuation risk value of each second public space model based on the passage hazard coefficient of the third public space model corresponding to each second public space model, the length of the evacuation route and the safety benchmark coefficient generated based on the shared impact coefficient of the shared second public space model.

[0078] Step S204: If the evacuation risk value is less than the preset evacuation risk threshold, set the ideal public safety behavior type to the ideal evacuation behavior type.

[0079] Step S205: If the evacuation risk value is greater than the preset evacuation risk threshold, set the ideal public safety behavior type to the ideal waiting-for-rescue behavior type.

[0080] In the aforementioned multimodal public safety behavior dynamic identification method, by quantitatively calculating the passage hazard coefficient and evacuation risk value, the feasibility of evacuation in various public spaces can be accurately assessed, providing a quantitative basis for the classification of ideal public safety behavior types. This makes the ideal public safety behavior types more in line with the actual situation and improves the rationality of public safety behavior guidance. By identifying ideal evacuation behavior types and ideal waiting-for-rescue behavior types, the appropriate ideal behavior type can be automatically selected based on the real-time risk status of the public space, ensuring that the ideal behavior type conforms to the current risk status. By identifying and analyzing the third public space model, evacuation routes can be planned according to the safety level of the target space. By comprehensively considering the safety status of the route and the target space, the safety of the guided evacuation behavior can be ensured, further improving the reliability of behavior guidance.

[0081] In an optional embodiment of this application, the environmental monitoring data may include video environmental monitoring data and smoke sensor monitoring data. Step S202, identifying the passage hazard coefficient of each third public space model in the first public space model based on the environmental monitoring data, may include:

[0082] Specifically, the public safety monitoring terminal can identify the obstacle distribution characteristics of the third public space model based on video environmental monitoring data, and generate a passage obstacle risk coefficient based on the obstacle distribution characteristics.

[0083] Optionally, the public safety monitoring terminal can input video environmental monitoring data into a convolutional neural network model to identify obstacle distribution features, obtain the obstacle distribution features of the third public space model, and generate a passage obstacle risk coefficient based on the obstacle distribution features.

[0084] Specifically, the public safety monitoring terminal can identify the smoke concentration characteristics and smoke type characteristics of the third public space model based on smoke sensor monitoring data, and generate a passage toxicity risk coefficient based on the smoke concentration characteristics and smoke type characteristics.

[0085] Optionally, the public safety monitoring terminal can set the smoke hazard concentration threshold to the corresponding smoke hazard concentration threshold for each smoke type based on the characteristics of the smoke type, and calculate the passage toxicity risk coefficient based on the smoke hazard concentration threshold and smoke concentration characteristics.

[0086] Specifically, the public safety monitoring terminal can calculate the passage risk coefficient by combining the passage obstacle risk coefficient and the passage toxicity risk coefficient.

[0087] Optionally, the public safety monitoring terminal can calculate the passage hazard coefficient by linearly weighting the passage obstacle hazard coefficient and the passage toxicity hazard coefficient.

[0088] Optionally, when the public safety monitoring terminal is equipped with a lightweight neural network model, the obstacle risk coefficient and the toxicity risk coefficient can be input into the lightweight neural network model to calculate the passage risk coefficient.

[0089] Furthermore, lightweight neural network models may include, but are not limited to, random forest models, convolutional neural networks, and inverse neural network models.

[0090] In the aforementioned multimodal dynamic identification method for public safety behaviors, by fusing and analyzing video modalities and smoke sensing modalities, it is possible to achieve multi-dimensional and all-round perception of the passage environment of the third public space model. This allows for the comprehensive capture of various safety hazards in the passage environment of the third public space model, providing reliable multimodal data support for determining ideal public safety behavior types. Through obstacle hazard coefficients and toxicity hazard coefficients, it is possible to achieve accurate quantitative assessment of passage hazards, distinguish the severity of different types of hazards, and thus accurately identify specific dangerous situations in the passage environment, improving the accuracy of passage risk control.

[0091] In an optional embodiment of this application, the expression for the evacuation risk value can be:

[0092]

[0093] In the formula, R i Let i be the evacuation risk value for the second public space model with serial number i. D represents the total number of third common spaces corresponding to the second common space model with index i. i,j I represents the passage hazard factor of the third public space with index j corresponding to the second public space model with index i. i,j L is the shared influence coefficient generated for the third public space with index j corresponding to the second public space model with index i, based on the shared second public space model. i,j S represents the length of the evacuation route corresponding to the third public space with index j, which is the second public space model with index i. i,j The safety benchmark coefficient is the third public space with index j corresponding to the second public space model with index i. The safety benchmark coefficient is used to characterize the completeness of the safety facilities in the third public space.

[0094] In an optional embodiment of this application, the actual public safety behavior type may include the actual waiting-for-rescue behavior type corresponding to the ideal waiting-for-rescue behavior type and the actual evacuation behavior type corresponding to the ideal evacuation behavior type. Please refer to [reference needed]. Figure 1 and Figure 2Step S103, based on personnel monitoring data and environmental monitoring data, and combined with the second public space model, identifies the actual public safety behavior types of the personnel corresponding to the personnel monitoring data, which may include:

[0095] Step S206: Based on environmental monitoring data and combined with the second public space model, identify the areas in the second public space model that require assistance.

[0096] Optionally, the public safety monitoring terminal can identify the hazard source based on environmental monitoring data, and can obtain the preset safe area in the second public space model, and combine the impact range of the hazard source and the safe area to identify the area to be rescued in the second public space model.

[0097] Step S207: Based on personnel monitoring data, combined with the personnel identification neural network model and the second public space model, the location information of the personnel is identified.

[0098] Optionally, the public safety monitoring terminal can input personnel monitoring data into a personnel recognition neural network model to identify personnel and generate their relative position information. The public safety monitoring terminal can then combine this relative position information with a second public space model to identify the personnel's position within the second public space model.

[0099] Step S208: If the location information of the person belongs to the area awaiting rescue, set the actual public safety behavior type to the actual awaiting rescue behavior type.

[0100] Step S209: If the location information of the personnel belongs to the fourth public space model outside the rescue area in the second public space model, set the actual public safety behavior type to the actual evacuation behavior type.

[0101] In the above-mentioned multimodal public safety behavior dynamic identification method, the personnel location monitoring can achieve high-precision tracking of the specific movements of personnel and improve the accuracy of identifying the actual public safety behavior type; by matching location information with spatial area, the actual public safety behavior type can be determined, which can achieve efficient identification of the actual behavior type of personnel, improve the real-time performance and accuracy of public safety behavior management, and thus improve the adaptability of behavior guidance information.

[0102] In an optional embodiment of this application, determining the conformity between the ideal public safety behavior type and the actual public safety behavior type to obtain the conformity determination result may include:

[0103] Specifically, if the actual public safety behavior type is the actual rescue-awaiting behavior type and the ideal public safety behavior type is the ideal rescue-awaiting behavior type, or if the actual public safety behavior type is the actual evacuation behavior type and the ideal public safety behavior type is the ideal evacuation behavior type, the public safety monitoring terminal can set the compliance judgment result to compliant.

[0104] Specifically, if the actual public safety behavior type is the actual evacuation behavior type and the ideal public safety behavior type is the ideal waiting-for-rescue behavior type, or if the actual public safety behavior type is the actual waiting-for-rescue behavior type and the ideal public safety behavior type is the ideal evacuation behavior type, the public safety monitoring terminal can set the compliance judgment result to deviation.

[0105] In an optional embodiment of this application, if the compliance determination result is compliant, public safety behavior execution guidance information is generated based on the ideal public safety behavior type, which may include:

[0106] Specifically, if the compliance judgment result is compliant and the ideal public safety behavior type is the ideal evacuation behavior type, the public safety monitoring terminal can use the personnel's location information as the starting point, take each of the second exits of the second public space model as the endpoint, and perform path planning based on the grid map in the second public space model to generate the second public space evacuation distance and the second optimal evacuation path of each of the second exits of the second public space model.

[0107] Specifically, the public safety monitoring terminal can generate a modified evacuation hazard coefficient for each second exit based on the evacuation distance of each second exit in the second public space model and the passage hazard coefficient of the corresponding third public space model for each second exit.

[0108] Specifically, the public safety monitoring terminal can set the second exit with the lowest evacuation risk coefficient as the optimal evacuation exit, and generate public safety action execution guidance information based on the optimal evacuation exit. This guidance information can be used to instruct personnel to evacuate to the optimal evacuation exit via a second optimal evacuation route, and then evacuate to the first exit of the first public space model based on a third public space model derived from the optimal evacuation exit.

[0109] In the aforementioned multimodal public safety behavior dynamic identification method, dynamic path planning can accurately quantify the distance of evacuation paths and generate optimal evacuation paths from the current location of personnel to each exit, making evacuation risk assessment more representative of the actual evacuation process and improving the scientific nature of evacuation path planning. By integrating distance and risk coefficient calculation, a comprehensive assessment of the overall risk of evacuation exits can be achieved, improving the accuracy of evacuation risk judgment. By setting the second exit with the lowest modified evacuation risk coefficient as the optimal evacuation exit, the optimal evacuation route can be quickly and accurately identified, reducing the risks during the evacuation process and improving the safety of evacuation behavior.

[0110] In one exemplary embodiment of this application, such as Figure 2 As shown, a multimodal public safety behavior dynamic recognition method is provided, including:

[0111] Step S101: Obtain multimodal monitoring data.

[0112] Step S102: Based on environmental monitoring data, the passage hazard coefficients of each third public space model in the first public space model are identified.

[0113] Step S103: Calculate the evacuation risk value of each second public space model based on the passage hazard coefficient of the third public space model corresponding to each second public space model.

[0114] Step S104: If the evacuation risk value is less than the preset evacuation risk threshold, set the ideal public safety behavior type to the ideal evacuation behavior type.

[0115] Step S105: If the evacuation risk value is greater than the preset evacuation risk threshold, set the ideal public safety behavior type to the ideal waiting-for-rescue behavior type.

[0116] Step S106: Based on environmental monitoring data and combined with the second public space model, identify the areas in the second public space model that require assistance.

[0117] Step S107: Based on personnel monitoring data, combined with the personnel identification neural network model and the second public space model, the location information of the personnel is identified.

[0118] Step S108: If the location information of the person belongs to the area awaiting rescue, set the actual public safety behavior type to the actual awaiting rescue behavior type.

[0119] Step S109: If the location information of the personnel belongs to the fourth public space model outside the rescue area in the second public space model, set the actual public safety behavior type to the actual evacuation behavior type.

[0120] Step S110: Determine the degree of conformity between the ideal public safety behavior type and the actual public safety behavior type, and obtain the degree of conformity judgment result.

[0121] Step S111: If the compliance judgment result is compliant, generate public safety behavior execution guidance information based on the ideal public safety behavior type.

[0122] Step S112: If the compliance judgment result is a deviation, generate public safety behavior change guidance information based on the actual public safety behavior type and the ideal public safety behavior type.

[0123] The aforementioned multimodal public safety behavior dynamic identification method can adjust behavioral guidance strategies according to real-time environmental changes, thereby improving the response capability to emergencies. Specifically, by acquiring multimodal monitoring data, it captures the environmental and personnel status in real time, calculates the passage hazard coefficient and evacuation risk value, and dynamically determines the ideal public safety behavior type, accurately adapting the behavioral guidance strategy to the real-time public safety environment. Simultaneously, by identifying the area awaiting rescue, determining the personnel location, and classifying the actual behavior type, it improves the clarity of the definition of the actual evacuation behavior type, ensuring the accuracy of evacuation behavior assessment. By judging the conformity between the ideal public safety behavior type and the actual evacuation behavior type, corresponding guidance information is generated, enabling flexible adjustment of behavioral guidance strategies to adapt to environmental changes. This allows for rapid and scientific guidance during emergencies, significantly improving the speed and effectiveness of public safety response.

[0124] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0125] Based on the same inventive concept, this application also provides a multimodal public safety behavior dynamic recognition system for implementing the multimodal public safety behavior dynamic recognition method described above. The solution provided by this system is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more embodiments of the multimodal public safety behavior dynamic recognition system provided below can be found in the limitations of the multimodal public safety behavior dynamic recognition method described above, and will not be repeated here.

[0126] In one exemplary embodiment, such as Figure 3 As shown, a multimodal public safety behavior dynamic recognition system 300 is provided, including:

[0127] The monitoring data acquisition module 301 can be used to acquire multimodal monitoring data, including personnel monitoring data and environmental monitoring data.

[0128] The ideal behavior analysis module 302 can be used to identify the ideal public safety behavior types of each second public space model in the first public space model based on environmental monitoring data and combined with the first public space model.

[0129] The actual behavior recognition module 303 can be used to identify the type of actual public safety behavior of personnel corresponding to the personnel monitoring data based on personnel monitoring data and environmental monitoring data, combined with the second public space model;

[0130] The conformity determination module 304 can be used to determine the conformity between the ideal public safety behavior type and the actual public safety behavior type, and obtain the conformity determination result.

[0131] The behavior execution guidance module 305 can be used to generate public safety behavior execution guidance information based on the ideal public safety behavior type if the compliance judgment result is compliant.

[0132] The behavior change guidance module 306 can be used to generate public safety behavior change guidance information based on the actual public safety behavior type and the ideal public safety behavior type if the compliance judgment result is a deviation. The public safety behavior change guidance information is used to guide personnel to change from performing the actual public safety behavior type to performing the ideal public safety behavior type.

[0133] In an optional embodiment of this application, the ideal behavior analysis module 302 can also be used for:

[0134] Based on environmental monitoring data, the passage hazard coefficients of each third public space model in the first public space model are identified. The third public space model is used to connect the first exit of the first public space model and the second exit of each second public space model.

[0135] The evacuation risk value of each second public space model is calculated based on the passage hazard coefficient of the third public space model corresponding to each second public space model.

[0136] If the evacuation risk value is less than the preset evacuation risk threshold, the ideal public safety behavior type will be set as the ideal evacuation behavior type.

[0137] If the evacuation risk value is greater than the preset evacuation risk threshold, the ideal public safety behavior type will be set to the ideal waiting-for-rescue behavior type.

[0138] In an optional embodiment of this application, the ideal behavior analysis module 302 can also be used for:

[0139] The obstacle distribution characteristics of the third public space model are identified based on video environmental monitoring data, and the obstacle risk coefficient is generated based on the obstacle distribution characteristics.

[0140] Based on smoke sensor monitoring data, the smoke concentration and smoke type characteristics of the third public space model are identified, and a traffic toxicity risk coefficient is generated based on the smoke concentration and smoke type characteristics.

[0141] The passage hazard coefficient is calculated by combining the passage obstacle hazard coefficient and the passage toxicity hazard coefficient.

[0142] In an optional embodiment of this application, the actual behavior recognition module 303 can also be used for:

[0143] Based on environmental monitoring data and combined with the second public space model, areas in the second public space model that require assistance are identified.

[0144] Based on personnel monitoring data, combined with a personnel identification neural network model and a second public space model, the location information of personnel is identified.

[0145] If a person's location information belongs to an area awaiting assistance, set the actual public safety behavior type to the actual awaiting assistance behavior type.

[0146] If the location information of a person belongs to the fourth public space model, which is outside the area awaiting rescue in the second public space model, the actual public safety behavior type will be set to the actual evacuation behavior type.

[0147] In an optional embodiment of this application, the behavior execution guidance module 305 can also be used for:

[0148] If the actual public safety behavior type is the actual rescue-awaiting behavior type and the ideal public safety behavior type is the ideal rescue-awaiting behavior type, or if the actual public safety behavior type is the actual evacuation behavior type and the ideal public safety behavior type is the ideal evacuation behavior type, the compliance judgment result will be set to compliant.

[0149] If the actual public safety behavior type is the actual evacuation behavior type and the ideal public safety behavior type is the ideal waiting-for-rescue behavior type, or if the actual public safety behavior type is the actual waiting-for-rescue behavior type and the ideal public safety behavior type is the ideal evacuation behavior type, the conformity judgment result will be set as deviation.

[0150] In an optional embodiment of this application, the behavior change guidance module 306 can also be used for:

[0151] If the compliance judgment result is compliant, and the ideal public safety behavior type is the ideal evacuation behavior type, the path planning is performed with the location information as the starting point and each second exit as the ending point, generating the second public space evacuation distance for each second exit.

[0152] Based on the evacuation distance of the second public space at each second exit and the passage hazard coefficient of the corresponding third public space model at each second exit, a modified evacuation hazard coefficient for each second exit is generated.

[0153] The second exit with the lowest evacuation risk coefficient is set as the optimal evacuation exit, and public safety action execution guidance information is generated based on the optimal evacuation exit.

[0154] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the multimodal public safety behavior dynamic recognition method as described above.

[0155] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0156] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0157] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A method for dynamic recognition of multimodal public safety behaviors, characterized in that, The method includes: Acquire multimodal monitoring data, which includes personnel monitoring data and environmental monitoring data; Based on the environmental monitoring data and combined with the first public space model, the ideal public safety behavior type of each second public space model in the first public space model is identified. Based on the personnel monitoring data and the environmental monitoring data, and in conjunction with the second public space model, the actual public safety behavior type of the personnel corresponding to the personnel monitoring data is identified. The conformity between the ideal public safety behavior type and the actual public safety behavior type is determined to obtain the conformity determination result. If the compliance judgment result is compliant, public safety behavior execution guidance information is generated based on the ideal public safety behavior type; If the conformity judgment result is a deviation, public safety behavior change guidance information is generated based on the actual public safety behavior type and the ideal public safety behavior type. The public safety behavior change guidance information is used to guide the personnel to change from performing the actual public safety behavior type to performing the ideal public safety behavior type.

2. The method according to claim 1, characterized in that, The ideal public safety behavior types include ideal waiting-for-rescue behavior types and ideal evacuation behavior types. Based on the environmental monitoring data and combined with the first public space model, the ideal public safety behavior types of each second public space model within the first public space model are identified, including: Based on the environmental monitoring data, the passage hazard coefficient of each third public space model in the first public space model is identified. The third public space model is used to connect the first exit of the first public space model and the second exit of each of the second public space models. The evacuation risk value of each second public space model is calculated based on the passage hazard coefficient of the third public space model corresponding to each second public space model; If the evacuation risk value is less than the preset evacuation risk threshold, the ideal public safety behavior type is set as the ideal evacuation behavior type; If the evacuation risk value is greater than the preset evacuation risk threshold, the ideal public safety behavior type is set as the ideal rescue-awaiting behavior type.

3. The method according to claim 2, characterized in that, The environmental monitoring data includes video environmental monitoring data and smoke sensor monitoring data. The step of identifying the access hazard coefficients of each third public space model in the first public space model based on the environmental monitoring data includes: Based on the video environment monitoring data, the obstacle distribution characteristics of the third public space model are identified, and a passage obstacle risk coefficient is generated according to the obstacle distribution characteristics. Based on the smoke sensor monitoring data, the smoke concentration characteristics and smoke type characteristics of the third public space model are identified, and a passage toxicity risk coefficient is generated based on the smoke concentration characteristics and smoke type characteristics. The passage hazard coefficient is calculated by combining the passage obstacle hazard coefficient and the passage toxicity hazard coefficient.

4. The method according to claim 2, characterized in that, The expression for the evacuation risk value is: In the formula, R i The evacuation risk value is for the second public space model with serial number i. D represents the total number of the third public spaces corresponding to the second public space model with index i. i,j I represents the passage hazard factor of the third public space with index j corresponding to the second public space model with index i. i,j L is the shared influence coefficient generated based on the shared second public space model for the third public space corresponding to the second public space model with index i and index j. i,j S is the length of the evacuation route corresponding to the third public space, which corresponds to the second public space model with index i and index j. i,j The safety benchmark coefficient is the safety reference coefficient of the third public space corresponding to the second public space model with serial number i and serial number j. The safety benchmark coefficient is used to characterize the completeness of the safety facilities of the third public space.

5. The method according to claim 2, characterized in that, The actual public safety behavior types include actual rescue-awaiting behavior types corresponding to the ideal rescue-awaiting behavior types and actual evacuation behavior types corresponding to the ideal evacuation behavior types. The step of identifying the actual public safety behavior types of personnel corresponding to the personnel monitoring data based on the personnel monitoring data and the environmental monitoring data, combined with the second public space model, includes: Based on the environmental monitoring data and in conjunction with the second public space model, identify the areas in the second public space model that require assistance. Based on the personnel monitoring data, combined with the personnel identification neural network model and the second public space model, the location information of the personnel is identified; If the location information of the person belongs to the area awaiting assistance, the actual public safety behavior type is set to the actual behavior awaiting assistance type; If the location information of the person belongs to the fourth public space model outside the area to be rescued in the second public space model, the actual public safety behavior type is set to the actual evacuation behavior type.

6. The method according to claim 5, characterized in that, The process of determining the conformity between the ideal public safety behavior type and the actual public safety behavior type to obtain the conformity determination result includes: If the actual public safety behavior type is the actual rescue-awaiting behavior type and the ideal public safety behavior type is the ideal rescue-awaiting behavior type, or if the actual public safety behavior type is the actual evacuation behavior type and the ideal public safety behavior type is the ideal evacuation behavior type, then the compliance judgment result is set to "compliant". If the actual public safety behavior type is the actual evacuation behavior type and the ideal public safety behavior type is the ideal waiting-for-rescue behavior type, or if the actual public safety behavior type is the actual waiting-for-rescue behavior type and the ideal public safety behavior type is the ideal evacuation behavior type, then the conformity judgment result is set to the deviation.

7. The method according to claim 5, characterized in that, If the compliance determination result is compliant, public safety behavior execution guidance information is generated based on the ideal public safety behavior type, including: If the conformity judgment result is conforming, and the ideal public safety behavior type is the ideal evacuation behavior type, take the location information as the starting point and each of the second exits as the ending point, perform path planning, and generate the second public space evacuation distance for each of the second exits; Based on the evacuation distance of the second public space at each second exit and the passage hazard coefficient of the third public space model corresponding to each second exit, the modified evacuation hazard coefficient of each second exit is calculated. The second exit with the lowest modified evacuation risk coefficient is set as the optimal evacuation exit, and the public safety action execution guidance information is generated based on the optimal evacuation exit.

8. A multimodal public safety behavior dynamic recognition system, characterized in that, The system includes: The monitoring data acquisition module is used to acquire multimodal monitoring data, which includes personnel monitoring data and environmental monitoring data. The ideal behavior analysis module is used to identify the ideal public safety behavior types of each second public space model in the first public space model based on the environmental monitoring data and in combination with the first public space model. The actual behavior recognition module is used to identify the actual public safety behavior type of the personnel corresponding to the personnel monitoring data based on the personnel monitoring data and the environmental monitoring data, combined with the second public space model; The conformity determination module is used to determine the conformity between the ideal public safety behavior type and the actual public safety behavior type, and obtain the conformity determination result. The behavior execution guidance module is used to generate public safety behavior execution guidance information based on the ideal public safety behavior type if the compliance judgment result is compliant. The behavior change guidance module is used to generate public safety behavior change guidance information based on the actual public safety behavior type and the ideal public safety behavior type if the compliance judgment result is a deviation. The public safety behavior change guidance information is used to guide the personnel to change from performing the actual public safety behavior type to performing the ideal public safety behavior type.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

Images