Police cloud intelligent management integrated platform based on multi-mode perception and intelligent decision

Abstract

The invention discloses a police cloud intelligent management integrated platform based on multi-modal perception and intelligent decision, and relates to the technical field of police cloud management, and the technical scheme is characterized in that the platform comprises an acquisition module for acquiring area basic data and perception data of a target monitoring area; processing the regional basic data to obtain a normal security factor, a factor decision weight, an early warning feature factor and an early warning diagnosis factor of the target monitoring region; processing the sensing data to obtain a security sensing factor of the target monitoring area; performing matching analysis on the security sensing factor and a normal security factor to obtain an abnormal sensing factor and an abnormal factor parameter of the target monitoring area; obtaining an abnormal influence level of the target monitoring area according to the factor decision-making weight and the abnormal factor parameter, and carrying out evaluation level determination according to the abnormal influence level and the area basic data to obtain an area evaluation level of the target monitoring area; the method can adapt to scene requirements, and can continuously improve the overall efficiency of a security and protection system.

Description

Technical Field

[0001] The present invention relates to the field of police cloud management technology, and more specifically, to an integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making. Background Art

[0002] In modern society, public safety and security management demands are becoming increasingly complex and diverse, and traditional security management models are gradually exposing significant shortcomings. From a security threat perspective, urban areas face frequent population mobility and increasingly complex business models. Traditional integrated police cloud intelligent management platforms offer a relatively simplistic approach to identifying and responding to security risks, focusing solely on simple threshold-based alarm mechanisms. For example, in crowded places, these platforms simply trigger an alarm when the number of people in the area exceeds a fixed threshold. This ignores the dynamics of human flow, the spatial characteristics of human distribution, and the differences in appropriate human flow thresholds across time periods and activity scenarios. Furthermore, they lack effective data fusion and mining capabilities, making it difficult to identify potential correlations within complex data or identify risk trends from historical data. When anomalies occur, they are unable to predict potential risks from multiple data sources, making it impossible to issue early warnings and implement intervention measures. Alarms are triggered only after an anomaly exceeds a simple threshold, leaving security management stuck in the post-event processing phase, easily missing the optimal opportunity for action. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the purpose of the present invention is to provide an integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making, characterized by comprising: Acquisition module: acquires regional basic data and perception data of the target monitoring area; The first processing module processes the regional basic data to obtain the normal security factors, factor decision weights, early warning characteristic factors and early warning diagnosis factors of the target monitoring area; The second processing module processes the perception data to obtain the security perception factor of the target monitoring area; Matching module: matches and analyzes the security perception factors with the normal security factors to obtain the abnormal perception factors and abnormal factor parameters of the target monitoring area; Evaluation module: obtains the abnormal impact level of the target monitoring area according to the factor decision weight and abnormal factor parameters, and determines the regional evaluation level of the target monitoring area based on the abnormal impact level and regional basic data; The third processing module processes the warning characteristic factors to obtain the target warning diagnosis factors; and obtains the security warning status of the target monitoring area based on the warning characteristic factors and the target warning diagnosis factors; Decision-making module: Obtain the security alarm cloud decision results of the target monitoring area based on the regional assessment level and security warning status.

[0005] Preferably, processing the regional basic data to obtain the normal security factor, factor decision weight, early warning characteristic factor and early warning diagnosis factor of the target monitoring area specifically includes the following steps: Obtain normal security data of the target monitoring area based on regional basic data; Processing the normal security data to obtain the normal security factor corresponding to the target monitoring area and the normal factor parameter interval corresponding to the normal security factor; Set the factor decision weights of the target monitoring area corresponding to the normal security factors based on the regional basic data; Acquire corresponding historical security anomaly data based on regional basic data, wherein the historical security anomaly data includes security anomaly perception data and security diagnosis results corresponding to the security anomaly perception data; Processing the security diagnosis results to obtain early warning diagnosis factors; The security anomaly perception data is processed to obtain early warning characteristic factors.

[0006] Preferably, processing the security anomaly sensing data to obtain the early warning characteristic factor specifically includes the following steps: Extract security features from security anomaly sensing data to obtain security anomaly features corresponding to the target monitoring area; Generate early warning feature factors based on security anomaly characteristics.

[0007] Preferably, processing the normal security data to obtain the normal security factor corresponding to the target monitoring area and the normal factor parameter interval corresponding to the normal security factor specifically includes the following steps: Extract security features based on normal security data to obtain normal security features corresponding to the target monitoring area; And based on the normal security characteristics, the normal security factors corresponding to the target monitoring area and the normal factor parameter intervals corresponding to the normal security factors are generated.

[0008] Preferably, the security diagnosis results are processed to obtain early warning diagnosis factors, specifically including the following steps: Perform security feature extraction on the security diagnosis results to obtain the security diagnosis features corresponding to the target monitoring area; Generate early warning diagnosis factors based on security diagnosis characteristics.

[0009] Preferably, processing the perception data to obtain the security perception factor of the target monitoring area specifically includes the following steps: Extract security features from the perception data to obtain security perception features of the target monitoring area; The security perception factor of the target monitoring area and the perception factor parameters corresponding to the security perception factor are generated based on the security perception characteristics.

[0010] Preferably, matching and analyzing the security perception factor with the normal security factor to obtain the abnormal perception factor and abnormal factor parameters of the target monitoring area specifically includes the following steps: Comparing the perception factor parameters of the security perception factor with the normal factor parameter interval of the normal security factor to obtain the abnormal perception factor; Get the interval center parameters of the normal factor parameter interval; The abnormal factor parameter of the abnormal perception factor is obtained according to the interval center parameter and the perception factor parameter.

[0011] Preferably, the perception factor parameter of the security perception factor is compared with the normal factor parameter interval of the normal security factor to obtain the abnormal perception factor, specifically: If the perception factor parameter of the security perception factor is not within the normal factor parameter interval of the normal security factor, the security perception factor is marked as an abnormal perception factor.

[0012] Preferably, processing the warning characteristic factors to obtain target warning diagnostic factors specifically includes the following steps: Obtain the target factor association path of the early warning characteristic factor; The target warning diagnostic factor corresponding to the warning characteristic factor is obtained according to the target factor association path.

[0013] Preferably, obtaining the security warning status of the target monitoring area according to the warning characteristic factor and the target warning diagnosis factor specifically includes the following steps: Obtaining security diagnostic features corresponding to the target early warning diagnostic factors; The security warning status of the target monitoring area is obtained based on the warning characteristic factors and security diagnosis characteristics.

[0014] Compared with the prior art, the present invention has the following beneficial effects: The present invention collects regional basic data and perception data through an acquisition module, including regional static features and real-time dynamic information. The first processing module and the second processing module process the two types of data respectively, thereby extracting normal security factors and security perception factors. The matching module can comprehensively identify abnormal security conditions by comparing the two to identify abnormal perception factors and abnormal factor parameters, and can capture abnormal gatherings of people, hidden dangers of equipment failures, and abnormal fluctuations in environmental parameters. The evaluation module introduces factor decision weights and determines the abnormal impact level in combination with abnormal factor parameters; then integrates regional basic data to determine the regional assessment level, and corresponds different levels to different risk levels and handling priorities. For high-impact abnormalities, police forces are deployed first and emergency plans are activated; for low-impact abnormalities, regular inspections are used. Through hierarchical management, resource waste and excessive response are avoided, and resource allocation efficiency is greatly improved. The decision module outputs customized alarm cloud decision results based on the regional assessment level and warning status. This application can adapt to scene requirements and continuously improve the overall effectiveness of the security system. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The module diagram of the integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making proposed by the present invention; Figure 2 This is a schematic diagram of the steps for calculating the perception factor parameters in the integrated police-cloud intelligent management platform based on multimodal perception and intelligent decision-making proposed by the present invention; Figure 3 The present invention proposes a schematic diagram of the steps for calculating abnormal factor parameters in the integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making. DETAILED DESCRIPTION

[0016] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0017] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0018] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it constitute a separate or selective embodiment that is mutually exclusive with other embodiments.

[0019] It should be noted in advance that the acquisition and processing of all images, information and data in the present invention are carried out under the premise of complying with relevant data protection laws and policies and obtaining authorization from the owners of the corresponding devices.

[0020] Reference Figure 1-Figure 3 shown.

[0021] The embodiment further illustrates the integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making proposed in the present invention.

[0022] The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making includes: Acquisition module: acquires regional basic data and perception data of the target monitoring area; The first processing module processes the regional basic data to obtain the normal security factors, factor decision weights, early warning characteristic factors and early warning diagnosis factors of the target monitoring area; The second processing module processes the perception data to obtain the security perception factor of the target monitoring area; Matching module: matches and analyzes the security perception factors with the normal security factors to obtain the abnormal perception factors and abnormal factor parameters of the target monitoring area; Evaluation module: obtains the abnormal impact level of the target monitoring area according to the factor decision weight and abnormal factor parameters, and determines the regional evaluation level of the target monitoring area based on the abnormal impact level and regional basic data; The third processing module processes the warning characteristic factors to obtain the target warning diagnosis factors; and obtains the security warning status of the target monitoring area based on the warning characteristic factors and the target warning diagnosis factors; Decision-making module: Obtain the security alarm cloud decision results of the target monitoring area based on the regional assessment level and security warning status.

[0023] The acquisition module obtains the regional basic data and perception data of the target monitoring area. The regional basic data includes the geographical environment of the area, the configuration of inherent security facilities and historical security event records; the perception data is real-time dynamic information collected by perception devices, such as personnel flow, changes in environmental parameters and equipment operating status. Perception devices include cameras and sensors.

[0024] The first processing module processes the basic regional data to obtain normal security factors. Normal security factors are a set of features extracted based on the long-term stable security status of the region. They represent the security operation status of the region under normal circumstances, such as the range of personnel density. At the same time, factor decision weights are set for different normal security factors based on the importance and management priority of the region, clarifying the influence weight of each factor in the subsequent decision-making process. Early warning feature factors and early warning diagnostic factors are mined from the basic data. Early warning feature factors focus on potential features that cause security anomalies. Early warning diagnostic factors are information formed by sorting and refining the diagnostic results of historical security anomalies. They are used to subsequently identify whether anomalies have occurred in the current scenario and determine the nature and category of the anomaly.

[0025] The second processing module extracts security perception factors from the real-time sensor data. These factors reflect the current security status of the target monitored area. The matching module compares the security perception factors with normal security factors. By analyzing the differences in parameters and characteristics between the perception factors and normal factors, it identifies abnormal perception factors—factors that deviate from the normal security pattern—and simultaneously obtains abnormal factor parameters. These abnormal factor parameters quantify the degree and characteristics of the abnormality, clarifying the deviation of the regional security status from the normal pattern.

[0026] The assessment module combines factor decision weights and anomaly factor parameters to determine the anomaly impact level. By incorporating regional basic data and comprehensively considering the area's scale, function, and security performance, the anomaly impact level is adjusted to determine the regional assessment level for the target monitoring area, thereby comprehensively evaluating the overall impact of the anomaly on the region.

[0027] The factor decision weight corresponding to each abnormal perception factor and the abnormal factor parameter are weighted and summed to obtain the abnormal impact level of the target monitoring area.

[0028] First, a preliminary scope is defined based on the anomaly impact level, then revised based on regional baseline data. For example, if the anomaly impact level is moderate, but the area is densely populated with limited escape routes, the regional assessment level will be increased. Conversely, if the area has abundant security resources and strong anomaly response capabilities, the regional assessment level will be relatively low, even if the anomaly impact level is moderate. A comprehensive assessment of the actual threat posed by the anomaly impact in a specific regional scenario is conducted, ultimately determining a regional assessment level that best reflects the actual conditions of the target monitoring area.

[0029] The third processing module processes the warning characteristic factors to obtain the target warning diagnosis factors required for current analysis. The security warning status of the target monitoring area is obtained by combining the warning characteristic factors and the target warning diagnosis factors.

[0030] The security cloud decision results for the target monitoring area are obtained based on the regional assessment level and security warning status. Based on the pre-set decision database, the assessment results and warning status of different levels are used to formulate security cloud decision results that are adapted to the current security needs of the target monitoring area. This includes police deployment strategies, security measure adjustment plans, and emergency response processes, providing execution guidance for actual security management actions and achieving security management of the target monitoring area.

[0031] The basic regional data is processed to obtain the normal security factors, factor decision weights, early warning characteristic factors, and early warning diagnosis factors of the target monitoring area, specifically including the following steps: Obtain normal security data of the target monitoring area based on regional basic data; Processing the normal security data to obtain the normal security factor corresponding to the target monitoring area and the normal factor parameter interval corresponding to the normal security factor; Set the factor decision weights of the target monitoring area corresponding to the normal security factors based on the regional basic data; Obtain corresponding historical security anomaly data based on regional basic data. The historical security anomaly data includes security anomaly perception data and security diagnosis results corresponding to the security anomaly perception data. Processing the security diagnosis results to obtain early warning diagnosis factors; The security anomaly perception data is processed to obtain early warning characteristic factors.

[0032] This application uses regional basic data to filter out normal security data generated by the target monitoring area under normal and stable security conditions, including daily personnel flow patterns, security equipment normal operation parameters, and environmental safety indicators. Personnel flow patterns include the number of people entering and exiting entrances during different hours of the weekday, security equipment normal operation parameters include the camera's trouble-free operating time and the threshold range of fire sensors, and environmental safety indicators include the area's temperature, humidity, and lighting intensity.

[0033] Data mining and feature engineering techniques are used to extract normal security factors from normal security data that reflect the essential laws of regional security, such as the density of people in core corridors during the morning rush hour and the average power consumption of security equipment at night. By calculating the mean and standard deviation of each factor, the normal factor parameter range is defined, clarifying the quantitative boundaries of normality. For example, the density of people in core corridors during the morning rush hour is 30-80 people / channel. This clearly defines the numerical range for safe and stable operation, providing a reference standard for subsequent identification of abnormal conditions. If real-time data exceeds this range, a security risk is initially determined.

[0034] Factor decision weights are designed to differentiate the impact of different normal security factors on regional security decisions. Each normal security factor is assigned a corresponding weight based on the regional attributes and management requirements included in the regional basic data. The regional attributes include whether the area is a densely populated commercial complex or a warehouse park for material storage. Management requirements include prioritizing personnel safety, material safety, and facility safety. In highly densely populated commercial plazas, the decision weight for the crowd congestion factor is higher than that for the equipment temperature factor, as abnormal crowd gatherings have a more significant direct impact on public safety. In warehouses storing flammable materials, the firefighting facility operation factor carries a higher weight, as ensuring a safe material storage environment is the core of management.

[0035] Decision weights are optimized and adjusted based on dynamic changes in regional baseline data, such as security facility upgrades and shifts in management priorities. For example, if intelligent facial recognition access control is added to a target surveillance area, the importance of the personnel identity verification factor increases, and the decision weight is adjusted accordingly. This ensures that factor weights are continuously aligned with actual security management needs, allowing the influence of each factor in subsequent decision-making to be more context-sensitive, thereby improving the scientific nature of decision-making.

[0036] Security anomaly perception data generated during historical security anomalies and the corresponding security diagnosis results are screened from regional basic data. Security anomaly perception data is the raw information collected by sensing devices during the anomaly process, such as key frames of video footage from sudden crowd gatherings, abnormal readings uploaded by sensors during equipment failures (such as current surges and temperature excursions), and access control alarm records triggered by illegal intrusions. This type of data carries the characteristics of the scene in which the anomaly occurred. Security diagnosis results include post-analysis reports and disposal records of the anomaly, clarifying the cause of the anomaly, its development process, disposal plan, and diagnostic results. The cause of the anomaly includes the lack of early warning of a promotional event, the development process includes the evolution from localized gatherings to channel congestion, and the disposal plan and results include the method of police deployment and the time required to evacuate the congestion.

[0037] Security diagnostic results are processed to generate early warning diagnostic factors; security anomaly perception data is processed to generate early warning characteristic factors. Semantic parsing and feature extraction are performed on security diagnostic results to generate early warning diagnostic factors. For example, the risk of crowd gathering caused by promotional activities and equipment failure caused by aging serve as the basis for rapid matching and judgment during subsequent anomaly diagnosis. Anomaly detection and feature engineering methods are used to extract key scene features that trigger anomalies from security anomaly perception data, thereby generating early warning characteristic factors. For example, if the channel occupant density increases by 200% within 10 minutes and the equipment temperature suddenly rises by 15°C within 5 minutes, if the real-time data matches these features during the subsequent real-time monitoring process, potential anomalies can be identified in advance, triggering the early warning process.

[0038] The parameter range of normal security factors is compared with the abnormal perception characteristics in historical abnormal data. If the real-time collected data exceeds the normal parameter range, it is quickly associated with the historical abnormal case library and matched with the corresponding warning feature factors and warning diagnosis factors to determine the type of current abnormality and the consequences caused. Abnormal types include crowd gathering and equipment failure, and the consequences caused include local congestion and large-scale power outages.

[0039] Factor decision weights play a regulatory role. When factors with high decision weights exhibit abnormalities, the overall risk assessment level is elevated, triggering a more urgent response process. For example, if the decision weight for a crowd congestion factor is high and its parameters are outside the normal range, even if other factors are normal, the risk is assessed as high, prioritizing police deployment and initiating emergency plans. As regional baseline data is continuously updated, the parameter ranges, decision weights, and early warning and diagnostic factors of normal security factors are iteratively optimized, driving the efficient operation of the integrated police cloud intelligent management platform.

[0040] Processing the security anomaly perception data to obtain warning characteristic factors specifically includes the following steps: Extract security features from security anomaly sensing data to obtain security anomaly features corresponding to the target monitoring area; Generate early warning feature factors based on security anomaly characteristics.

[0041] Security anomaly perception data includes abnormal images captured by surveillance cameras, abnormal readings from various sensors, and abnormal access recorded by the access control system. Abnormal images captured by cameras include sudden crowds and unusual object movement. Abnormal readings from sensors include excessive smoke concentrations and sudden temperature increases. Abnormal access recorded by the access control system includes frequent attempts by unauthorized persons to enter.

[0042] For video data, it identifies changes in key elements within the image, such as detecting rapid changes in crowd density over a short period of time and marking them as abnormal crowd gathering features. It also determines whether an object's trajectory deviates from its normal path and extracts the characteristics of its abnormal movement trajectory. Sensor data is compared with historical normal data to identify significant deviations, such as a sudden jump in smoke concentration from the normal 0-5ppm to 20ppm, and extracts the characteristics of an abnormal smoke concentration peak. This processing allows for the precise extraction of security anomaly features from security anomaly perception data, thereby outlining the key manifestations of security anomalies in the target monitoring area.

[0043] After extracting anomaly security features, they must be converted into early warning feature factors that can be used for intelligent decision-making. Classification and quantification are performed based on feature type, severity, and impact range. For abnormal crowd gathering features, the density and proportion of the crowd gathering area are calculated in conjunction with regional basic data, which is then converted into a crowd gathering early warning factor to clarify its impact on regional security risks. Equipment overheating early warning factors are generated based on information such as sudden temperature rise characteristics, associated equipment type, and environmental conditions.

[0044] Processing the normal security data to obtain the normal security factor corresponding to the target monitoring area and the normal factor parameter interval corresponding to the normal security factor specifically includes the following steps: Extract security features based on normal security data to obtain normal security features corresponding to the target monitoring area; And based on the normal security characteristics, the normal security factors corresponding to the target monitoring area and the normal factor parameter intervals corresponding to the normal security factors are generated.

[0045] Key information representative of the stable operational status of regional security can be extracted from normal security data. For example, traffic flow patterns across different time periods and regions can be determined using personnel flow data. Features such as average travel time per person in core corridors during the morning rush hour on weekdays and average regional traffic density during off-peak hours can be extracted. Equipment operation data can be used to analyze the voltage, current, and temperature fluctuations during normal operation, and to summarize features such as the average daily online duration of security cameras and the normal response delay range of fire sensors. These extracted features collectively constitute the normal security features corresponding to the target monitoring area. These extracted features are then used to generate a normal security factor and its corresponding normal factor parameter range. The normal security factor transforms normal security features into core indicators useful for security analysis and decision-making. For example, personnel flow-related features such as average travel time per person in core corridors during the morning rush hour on weekdays and average regional traffic density during off-peak hours can be integrated into a regional dynamic traffic distribution factor. Equipment operation features such as the average daily online duration of security cameras and the normal response delay range of fire sensors can be summarized into a security equipment steady-state operation factor. The corresponding normal factor parameter range is determined for each normal security factor. By calculating the mean, standard deviation and extreme value distribution of characteristic data, the numerical range of the factor during safe and stable operation is clarified. For example, the parameter range corresponding to the regional pedestrian dynamic distribution factor is defined as the average travel time per person in the core channel during the morning rush hour on weekdays is 15-30 seconds, and the average regional pedestrian density during non-peak hours is 5-15 people / 100 square meters. This range clearly defines the fluctuation range of the factor under normal security conditions. If subsequent real-time monitoring data exceeds this range, it is determined that the regional security status is abnormal.

[0046] The security diagnosis results are processed to obtain early warning diagnosis factors, which specifically includes the following steps: Perform security feature extraction on the security diagnosis results to obtain the security diagnosis features corresponding to the target monitoring area; Generate early warning diagnosis factors based on security diagnosis characteristics.

[0047] Security diagnosis results include the cause of the anomaly, its development process, and the conclusion of the resolution. For example, a diagnostic report on a fire warning delay caused by aging equipment can extract the core elements related to equipment aging, warning delay, and fire risk, and then transform them into structured security diagnosis features, such as those caused by functional anomalies due to equipment operation and maintenance deficiencies and security response chain failures.

[0048] Generate early warning diagnostic factors based on the extracted security diagnostic features. Convert diagnostic features into quantitative indicators for security decision-making. Classify and assign values ​​based on the risk level, scope of impact, and degree of correlation of the features. For functional abnormalities caused by the lack of equipment operation and maintenance, combine them with regional basic data to convert them into equipment operation and maintenance health factors, and use historical data to train the model to determine the risk threshold corresponding to the factor. For example, if the operation and maintenance missing rate exceeds 30%, an early warning is triggered. At the same time, the handling effect of past abnormal events is associated with the factor to give the early warning weight, so that the early warning diagnostic factor can not only identify abnormal patterns, but also predict the consequences of risks. For example, the fire risk correlation feature is converted into a fire warning correlation factor. If the real-time data matches the factor, the historical fire handling plan is quickly associated.

[0049] The sensing data is processed to obtain the security sensing factor of the target monitoring area, which specifically includes the following steps: Extract security features from the perception data to obtain security perception features of the target monitoring area; The security perception factor of the target monitoring area and the perception factor parameters corresponding to the security perception factor are generated based on the security perception characteristics.

[0050] Security feature extraction from sensor data yields security perception features for the target surveillance area. Video data identifies human behavior and object states within the footage. Human behavior includes gatherings and running, while object states include piled items and displaced equipment. This identifies features of abnormal human gatherings and illegal object movement. Sensor data identifies fluctuations in values, such as sudden temperature rises and excessive smoke concentrations, to identify abnormal environmental parameter features and equipment failure warnings. Access control data is analyzed to identify matching access frequencies with personnel permissions, yielding features of abnormal personnel access and warnings of overreaching permissions. These extracted features collectively constitute the security perception features for the target surveillance area.

[0051] Based on these security perception features, security perception factors and corresponding perception factor parameters are generated. For example, the personnel-related features of abnormal personnel gatherings and abnormal personnel passages are integrated into a regional personnel dynamic security factor; the equipment environment-related features of abnormal environmental parameters and equipment failure warning features are summarized into a regional facility environment security factor. During the factor generation process, the perception factor parameters corresponding to each security perception factor are determined by combining the quantification of perception features. The numerical representation of the factor in its current state is clarified by statistically analyzing the mean, fluctuation range, and duration of the characteristic data. For example, the parameters corresponding to the regional personnel dynamic security factor may include the number of abnormal gatherings and the duration of the gathering, while the parameters for the regional facility environment security factor include the environmental parameter exceeding the standard value and the duration of the equipment failure. These parameters quantify the state of the perception factor and will be subsequently compared with normal security factors, warning feature factors, etc., providing data support for identifying anomalies and assessing risks.

[0052] The security perception factor is matched with the normal security factor to obtain the abnormal perception factor and abnormal factor parameters of the target monitoring area, which specifically includes the following steps: Comparing the perception factor parameters of the security perception factor with the normal factor parameter interval of the normal security factor to obtain the abnormal perception factor; Get the interval center parameters of the normal factor parameter interval; The abnormal factor parameter of the abnormal perception factor is obtained according to the interval center parameter and the perception factor parameter.

[0053] The abnormal perception factor is obtained by comparing the perception factor parameters of the security perception factor with the normal factor parameter interval of the normal security factor, specifically: If the perception factor parameter of the security perception factor is not within the normal factor parameter interval of the normal security factor, the security perception factor is marked as an abnormal perception factor.

[0054] The security perception factor carries information about the current security status of the target monitoring area, and its corresponding perception factor parameter is a quantitative representation of the real-time status. The normal factor parameter range for the normal security factor is the safe operating boundary defined based on historical normal data. The two are compared one by one. When a perception factor parameter exceeds the normal factor parameter range, the corresponding security perception factor is determined to be an abnormal perception factor. For example, the normal factor parameter range specifies that the normal range for regional population density is 10-50 people / 100 square meters. If the real-time perceived regional population density parameter is 60 people / 100 square meters, the regional population density perception factor is marked as an abnormal perception factor, allowing for quick identification of deviations from the normal pattern for regional security status.

[0055] The interval center parameter refines the core characteristics of the normal factor parameter interval and is obtained by calculating the mean of the interval. For example, if the normal factor parameter interval is 10-50 people / 100 square meters, the interval center parameter is 30 people / 100 square meters. This parameter is used to measure the degree of abnormal deviation.

[0056] The abnormal factor parameter is calculated based on the interval center parameter and the perception factor parameter. The difference between the perception factor parameter and the interval center parameter is used to determine the specific value of the abnormal deviation. For example, if the perception factor parameter is 60 people / 100 square meters and the interval center parameter is 30 people / 100 square meters, the difference between the two is 30 people / 100 square meters. This value is the abnormal factor parameter of the abnormal perception factor, clearly showing the deviation of the current state from the normal core state.

[0057] The early warning characteristic factors are processed to obtain target early warning diagnostic factors, which specifically includes the following steps: Obtain the target factor association path of the early warning characteristic factor; The target warning diagnostic factor corresponding to the warning characteristic factor is obtained according to the target factor association path.

[0058] Early warning characteristic factors are key feature identifiers extracted from historical security anomaly perception data for identifying potential risks. Target factor association paths are pre-constructed logical links that describe the relationship between early warning characteristic factors and early warning diagnostic factors. Actively obtain the target factor association path corresponding to the early warning characteristic factor. This path is mined using machine learning methods based on a large amount of historical security data. It records which early warning diagnostic factors the early warning characteristic factor is typically associated with when it appears historically, as well as the triggering conditions and logical sequence of this association. For example, when there is an early warning characteristic factor that indicates a sharp increase in the density of people in a certain channel within a short period of time, its target factor association path will sort out the intermediate association nodes that the characteristic factor will sequentially associate with the risk of gathering people and the hidden danger of channel blockage, ultimately pointing to the stampede risk early warning diagnostic factor.

[0059] Specifically, the target factor association path is constructed through the following steps: Based on historical data, first-sequence data on warning characteristic factors and target warning diagnostic factors are obtained, the first-sequence data are split into multiple second-sequence data, the appearance time of each element in the second-sequence data is obtained, the second-sequence data are clustered based on the appearance time, and multiple clustering results are obtained. The warning diagnostic factors of the warning characteristic factors and the target factor association paths are determined according to the number of second-sequence data included in the clustering results.

[0060] For example, historical data includes temperature data collected by various sensors before an accident occurs, as well as changes in crowd density obtained through video footage. The above data are encoded, and the first sequence of data is constructed based on the time of occurrence. For example, the first sequence of data is ABCDE, where E is the accident that occurred, that is, the target early warning diagnostic factor, such as someone fainting. ABCD is the data that appeared before the accident, that is, the early warning characteristic factor, such as A represents a 200% increase in crowd density, and B represents a 2°C increase in indoor temperature.

[0061] The first sequence of data is then split to obtain the second sequence of data. For example, ABCDE can be split into AE, ABE, ABCE, and ABCDE. The purpose is to analyze the degree of correlation between each data point and the final accident. The occurrence time of each element in the second sequence of data is then obtained, such as A appears at 15:00:02 and B appears at 15:03:02. The time difference between elements is then obtained, such as the time difference between A and B, and the time difference between A and C. It should be noted that there is a large amount of accident data in historical data, so the above first sequence of data may contain multiple entries, such as ABCDF, where F represents another accident. In other words, the second sequence of data formed after the split will also include multiple sequences of different accidents.

[0062] Then, clustering is performed on each second series data, such as clustering the second time series AE. The time difference of the elements is used as the clustering feature during clustering. The clustering algorithm can use K-means or DBSCAN algorithm.

[0063] After clustering is completed, multiple clusters will be obtained. The number of second sequence data included in each cluster will be obtained, and the cluster with the largest number will be selected. The target factor association path will be determined based on the second sequence data included in it. If it is found through analysis that a certain cluster includes the most second sequence data, and the second sequence data included is ABE, then when the warning characteristic factors A and B appear, the target warning diagnostic factor E will be associated.

[0064] Based on the acquired target factor association path, the warning characteristic factors are matched and associated with the corresponding warning diagnostic factors along the established logical links. Because the path clearly defines how to associate the warning characteristic factors with specific warning diagnostic factors, it is possible to find the target warning diagnostic factor corresponding to the current warning characteristic factor. When the warning characteristic factor is detected in real-time monitoring, the corresponding target warning diagnostic factor can be quickly located through the association path, clarifying the specific security risk type and diagnostic conclusion that the warning characteristic factor may trigger, providing direct and critical diagnostic evidence for subsequent security warning status analysis and decision-making.

[0065] The security warning status of the target monitoring area is obtained based on the warning characteristic factors and the target warning diagnosis factors, which specifically includes the following steps: Obtain security diagnostic features corresponding to target early warning diagnostic factors; The security warning status of the target monitoring area is obtained based on the warning characteristic factors and security diagnosis characteristics.

[0066] The target early warning diagnostic factor is a core factor extracted from historical security anomaly data to predict potential risks. Its corresponding security diagnostic features are derived from the mining of diagnostic results of similar security anomalies in the past, including key information such as the patterns of anomaly occurrence, triggering conditions, and impact range. A data retrieval and matching mechanism is used to extract security diagnostic features associated with the target early warning diagnostic factor from the historical security diagnostic database. For example, when the target early warning diagnostic factor points to the risk of a stampede caused by a gathering of people, the corresponding security diagnostic features may include the crowd density threshold and abnormal crowd flow rate characteristics for a specific area.

[0067] The security warning status of the target monitoring area is determined based on warning characteristic factors and security diagnostic features. Warning characteristic factors are features extracted from the real-time sensor data of the current monitoring area that may indicate risks. These features are compared and integrated with the security diagnostic features to determine the degree of match and correlation strength. If the warning characteristic factors are highly consistent with the security diagnostic features—for example, if the warning characteristic factors extracted for population density and crowd flow rate in a certain area are consistent with the security diagnostic features corresponding to the risk of stampede caused by crowd gatherings—then the corresponding security risk is determined in that area, and the warning level and risk type are determined.

[0068] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0069] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making is characterized by: include: Acquisition module: acquires regional basic data and perception data of the target monitoring area; The first processing module processes the regional basic data to obtain the normal security factors, factor decision weights, early warning characteristic factors and early warning diagnosis factors of the target monitoring area; The second processing module processes the perception data to obtain the security perception factor of the target monitoring area; Matching module: matches and analyzes the security perception factors with the normal security factors to obtain the abnormal perception factors and abnormal factor parameters of the target monitoring area; Evaluation module: obtains the abnormal impact level of the target monitoring area according to the factor decision weight and abnormal factor parameters, and determines the regional evaluation level of the target monitoring area based on the abnormal impact level and regional basic data; The third processing module: processes the warning characteristic factors to obtain the target warning diagnosis factors; Obtain the security warning status of the target monitoring area based on the warning characteristic factors and target warning diagnosis factors; Decision-making module: Obtain the security alarm cloud decision results of the target monitoring area based on the regional assessment level and security warning status.

2. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making according to claim 1 is characterized in that: The basic regional data is processed to obtain the normal security factors, factor decision weights, early warning characteristic factors, and early warning diagnosis factors of the target monitoring area, specifically including the following steps: Obtain normal security data of the target monitoring area based on regional basic data; Processing the normal security data to obtain the normal security factor corresponding to the target monitoring area and the normal factor parameter interval corresponding to the normal security factor; Set the factor decision weights of the target monitoring area corresponding to the normal security factors based on the regional basic data; Acquire corresponding historical security anomaly data based on regional basic data, wherein the historical security anomaly data includes security anomaly perception data and security diagnosis results corresponding to the security anomaly perception data; Processing the security diagnosis results to obtain early warning diagnosis factors; The security anomaly perception data is processed to obtain early warning characteristic factors.

3. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making according to claim 2 is characterized in that: Processing the security anomaly perception data to obtain warning characteristic factors specifically includes the following steps: Extract security features from security anomaly sensing data to obtain security anomaly features corresponding to the target monitoring area; Generate early warning feature factors based on security anomaly characteristics.

4. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making according to claim 3 is characterized in that: Processing the normal security data to obtain the normal security factor corresponding to the target monitoring area and the normal factor parameter interval corresponding to the normal security factor specifically includes the following steps: Extract security features based on normal security data to obtain normal security features corresponding to the target monitoring area; And based on the normal security characteristics, the normal security factors corresponding to the target monitoring area and the normal factor parameter intervals corresponding to the normal security factors are generated.

5. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making according to claim 4 is characterized in that: The security diagnosis results are processed to obtain early warning diagnosis factors, which specifically includes the following steps: Perform security feature extraction on the security diagnosis results to obtain the security diagnosis features corresponding to the target monitoring area; Generate early warning diagnosis factors based on security diagnosis characteristics.

6. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making according to claim 1 is characterized in that: The sensing data is processed to obtain the security sensing factor of the target monitoring area, which specifically includes the following steps: Extract security features from the perception data to obtain security perception features of the target monitoring area; The security perception factor of the target monitoring area and the perception factor parameters corresponding to the security perception factor are generated based on the security perception characteristics.

7. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making according to claim 4 is characterized in that: The security perception factor is matched with the normal security factor to obtain the abnormal perception factor and abnormal factor parameters of the target monitoring area, which specifically includes the following steps: Comparing the perception factor parameters of the security perception factor with the normal factor parameter interval of the normal security factor to obtain the abnormal perception factor; Get the interval center parameters of the normal factor parameter interval; The abnormal factor parameter of the abnormal perception factor is obtained according to the interval center parameter and the perception factor parameter.

8. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making according to claim 7 is characterized in that: The abnormal perception factor is obtained by comparing the perception factor parameters of the security perception factor with the normal factor parameter interval of the normal security factor, specifically: If the perception factor parameter of the security perception factor is not within the normal factor parameter interval of the normal security factor, the security perception factor is marked as an abnormal perception factor.

9. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making according to claim 1 is characterized in that: The early warning characteristic factors are processed to obtain target early warning diagnostic factors, which specifically includes the following steps: Obtain the target factor association path of the early warning characteristic factor; The target warning diagnostic factor corresponding to the warning characteristic factor is obtained according to the target factor association path.

10. The integrated police cloud intelligent management platform based on multimodal perception and intelligent decision-making according to claim 1 is characterized in that: The security warning status of the target monitoring area is obtained based on the warning characteristic factors and the target warning diagnosis factors, which specifically includes the following steps: Obtaining security diagnostic features corresponding to the target early warning diagnostic factors; The security warning status of the target monitoring area is obtained based on the warning characteristic factors and security diagnosis characteristics.

Images