Risk area early warning information generation method for smart city and internet of things system

Abstract

The embodiment of the specification provides a risk area early warning information generation method and an Internet of Things system for a smart city. The method comprises the following steps: acquiring a query instruction of a user on each area; acquiring monitoring images of at least one target area based on the query instruction; processing the monitoring images to determine a personnel suspicious index of at least one person in the monitoring images; determining a suspicious person in response to the personnel suspicious index meeting a preset condition; determining distances between the suspicious persons based on monitoring images of a plurality of adjacent frames; determining suspicious groups corresponding to the suspicious persons based on the distances; determining a risk index of the target area based on a sum of suspicious group indexes of each suspicious group in the target area; and generating early warning information in response to the risk index of the target area being greater than a first threshold value.

Description

[0001] Divisional Statement

[0002] The present application is a divisional application of the Chinese application with the application number 202211194707.1, the application date of September 29, 2022, and the invention name of "a smart city regional public security management early warning method and system based on the Internet of Things". TECHNICAL FIELD

[0003] The present specification relates to the field of public security management, in particular to a risk area early warning information generation method for smart cities and an Internet of Things system. BACKGROUND

[0004] Public security management is the basic guarantee for the normal operation of people's social life. The existing public security management generally manages urban public security through real-time monitoring of monitoring information of communities, streets, shopping malls, parking lots and other places or arranging patrol personnel to patrol. When it is preliminarily judged that there may be suspicious behavior in the area through monitoring information, patrol personnel are immediately arranged to check the scene, which may cause insufficient allocation of police resources. Therefore, it is hoped to propose a risk area early warning information generation method for smart cities and an Internet of Things system, which can further determine the degree of public security work development to improve the efficiency of public security management. SUMMARY

[0005] One of the embodiments of the present specification provides a risk area early warning information generation method for smart cities, which is executed by a public security management platform, comprising: acquiring a query instruction of a user to each area; acquiring monitoring images of at least one target area based on the query instruction; processing the monitoring images to determine a personnel suspicious index of at least one person in the monitoring images; determining a suspicious person in response to the personnel suspicious index meeting a preset condition; determining the distance between the suspicious persons based on the monitoring images of multiple adjacent frames; determining the suspicious groups corresponding to the suspicious persons based on the distance; determining the risk index of the target area based on the sum of the suspicious group indexes of each suspicious group in the target area; and generating early warning information in response to the risk index of the target area being greater than a first threshold value.

[0006] One of the embodiments of the present specification also provides a risk area early warning information generation Internet of Things system for a smart city, the Internet of Things system comprises a user platform, a service platform, a public security management platform, a sensor network platform and an object platform which interact in sequence, the public security management platform is configured to perform the following operations: based on the user platform, acquiring a query instruction of a user to each area, the user platform is configured as at least one terminal device; in response to the query instruction, the public security management platform acquires monitoring images of at least one target area from at least one monitoring device of the at least one target area based on a sensor network sub-platform of the sensor network platform, the at least one monitoring device is configured in different object platforms; wherein the sensor network platform adopts different sensor network sub-platforms to store, process and / or transmit data of different object platforms, the sensor network sub-platform corresponds to different target areas; the public security management platform adopts different management sub-platforms to store, process and / or transmit data, and performs data aggregation, data processing and data transmission through a general platform of the public security management platform; based on the sensor network sub-platform, the monitoring images of the corresponding target areas are sent to the management sub-platform; based on the management sub-platform, the monitoring images are processed to determine a personnel suspicious index of at least one person in the monitoring images; in response to the personnel suspicious index meeting a preset condition, a suspicious person is determined; based on a plurality of adjacent frames of the monitoring images, a distance between the suspicious persons is determined; based on the distance, a suspicious group corresponding to the suspicious persons is determined; based on the sum of suspicious group indexes of each suspicious group in the target area, a risk index of the target area is determined; in response to the risk index of the target area being greater than a first threshold value, early warning information is generated.

[0007] One of the embodiments of the present specification also provides a risk area early warning information generation device for a smart city, comprising a processor, the processor is used to execute the risk area early warning information generation method for a smart city as described above.

[0008] One of the embodiments of the present specification also provides a computer readable storage medium, the storage medium stores computer instructions, when the computer reads the computer instructions, the computer executes the risk area early warning information generation method for a smart city as described above.

[0009] The beneficial effects of the embodiments of the present specification include at least: (1) the risk index of the target area can be determined in real time based on the monitoring image, improving the timeliness of the early warning; after the risk index exceeds a certain threshold, the early warning can be generated, and the unmanned aerial vehicle can be controlled to go to the corresponding target area for further monitoring, so as to manage the public security of the area and improve the management efficiency; (2) through processing of multiple risk indexes of the target area, the risk index of the target area can be determined in multiple dimensions and comprehensively. When the average value of the risk index is greater than a second threshold, the target area is determined as a daily patrol point, the monitoring intensity of the target area is increased, and the safety of the area is improved. In addition, by associating the patrol frequency with the proportion of suspicious groups with high suspicious indexes in the suspicious group distribution vector, the target area can be patrolled more targetedly, so that the patrol is more efficient, and the patrol resources are used reasonably; (3) by considering the correlation between different area types of the target area and the risk index of the target area, the risk index of the target area can be evaluated more targetedly by the public security management platform. In addition, by obtaining the stay duration of the suspicious person and determining the risk index of the target area by comprehensively monitoring the risk index of each person in the multiple persons in the image, the suspicious index of the person and the risk index of the target area can be accurately evaluated. BRIEF DESCRIPTION OF DRAWINGS

[0010] The present specification will be further illustrated in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, the same numbers represent the same structures, in which:

[0011] Figure 1 is an application scenario schematic diagram of a regional public security management early warning system according to some embodiments of the present specification;

[0012] Figure 2 is an exemplary block diagram of a regional public security management early warning system according to some embodiments of the present specification;

[0013] Figure 3 is an exemplary flowchart of an Internet of Things-based smart city regional public security management early warning method according to some embodiments of the present specification;

[0014] Figure 4 is an exemplary flowchart of a method for determining a risk index based on a suspicious person according to some embodiments of the present specification;

[0015] Figure 5 is an exemplary schematic diagram of determining a trajectory suspicious degree according to some embodiments of the present specification;

[0016] Figure 6 is an exemplary flowchart of a method for determining a risk index based on a suspicious group according to some embodiments of the present specification. DETAILED DESCRIPTION

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present specification, the drawings needed to be used in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some examples or embodiments of the present specification, and for those skilled in the art, the present specification can also be applied to other similar scenarios without creative labor on the basis of these drawings. Unless it is clear from the language environment or otherwise stated, the same reference numbers in the drawings represent the same structure or operation.

[0018] It should be understood that the "system", "device", "unit" and / or "module" used herein is a method for distinguishing different components, elements, parts, sections or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0019] Unless the context clearly indicates otherwise, the words "one", "an", "a", and / or "the" do not mean a single number, but can also include a plurality. Generally, the terms "comprise" and "include" only indicate that the steps and elements explicitly identified are included, and these steps and elements do not constitute an exclusive list, and the method or device can also include other steps or elements.

[0020] Flowcharts are used in the present specification to illustrate the operations performed by the system according to the embodiments of the present specification. It should be understood that the preceding or subsequent operations are not necessarily performed in sequence. On the contrary, each step can be processed in reverse order or simultaneously. At the same time, other operations can be added to these processes, or one or more steps of the operation can be removed from these processes.

[0021] Figure 1 is a schematic diagram of the application scenario of the regional public security management early warning system according to some embodiments of the present specification.

[0022] In some embodiments, the regional public security management early warning system 100 can include a processing device 110, a network 120, a terminal 130, a storage device 140, a monitoring device 150, a drone 160. In some embodiments, the regional public security management early warning system 100 can be used for public security management in public areas. The regional public security management early warning system 100 can realize the Internet of Things-based smart city regional public security management early warning by implementing the methods and / or processes disclosed in the present specification. In some embodiments, the components in the regional public security management early warning system 100 can be connected and / or communicated with each other via the network 120 (e.g., wireless connection, wired connection or a combination thereof). For example, the processing device 110 can be connected to the storage device 120 through the network 120.

[0023] The processing device 110 can process data and / or information obtained from other devices or system components. In some embodiments, the processing device 110 can access information and / or data from the terminal 130, the storage device 140, the monitoring device 150, and / or the drone 160. For example, the processing device 110 obtains monitoring images of a target area from the monitoring device 150. In some embodiments, the processing device 110 can process information and / or data obtained from the terminal 130, the storage device 140, the monitoring device 150, and / or the drone 160. For example, the processing device 110 can determine a risk index of a target area based on the monitoring images, and produce a warning message when the risk index is greater than a first threshold, and control the drone 160 to monitor the target area. In some embodiments, the processing device 110 can be a server or a group of servers. In some embodiments, one or more different servers can be configured as a public security management platform, a service platform. For example, the processing device 110 can include a first server and a second server, where the first server can be configured as a service platform, and the second server can be configured as a public security management platform.

[0024] The network 120 can include any suitable network that provides for the exchange of information and / or data among the various components of the IoT-based smart city regional public security management warning system 100. Information and / or data can be exchanged between one or more components of the IoT-based smart city regional public security management warning system 100 (e.g., the processing device 110, the terminal 130, the storage device 140, the monitoring device 150, and / or the drone 160) via the network 120. For example, the network 120 can send monitoring images of a target area obtained from the monitoring device 150 to the processing device 110. In some embodiments, the network 120 can be any one or more of a wired network or a wireless network. In some embodiments, the network can be various topologies or combinations of topologies, such as point-to-point, shared, hub-and-spoke, etc. In some embodiments, the network 120 can include one or more network access points. For example, the network 120 can include wired or wireless network access points.

[0025] The terminal 130 can communicate and / or connect with the processing device 110, the storage device 140, the monitoring device 150, and the drone 160. For example, the terminal 130 can send one or more control instructions to the monitoring device 150 through the network 120 to control the monitoring device 150 to take pictures of a target area according to the instructions, and / or send one or more control instructions to the drone 160 to control the drone 160 to conduct reconnaissance of the target area according to the instructions. For another example, a user can record suspicious persons in images taken by the monitoring device 150 through the terminal 130, and the terminal 130 can transmit the images to the storage device 140 for storage, or send the images to the processing device 110 for subsequent processing. In some embodiments, the user can be a police officer or a street manager. In some embodiments, the terminal 130 can be one or any combination of a mobile device 130-1, a tablet computer 130-2, a laptop computer 130-3, and other devices with input and / or output functions. The above examples are only used to illustrate the generality of the terminal 130 device and not to limit the scope thereof.

[0026] The storage device 140 can be used to store data, instructions, and / or any other information. In some embodiments, the storage device 140 can store data acquired from the processing device 110, the terminal 130, the monitoring device 150, and / or the drone 160. For example, the storage device 140 can be used to store monitoring images of target objects acquired by the monitoring device 150. In some embodiments, the storage device 140 can store data and / or instructions used by the processing device 110 to perform or use to complete the exemplary methods described in this specification.

[0027] In some embodiments, the storage device 140 can include a mass storage device, a removable storage device, a volatile read / write memory, a read-only memory (ROM), or any combination thereof. In some embodiments, the storage device 140 can be implemented on a cloud platform.

[0028] The monitoring device 150 can be used to collect data and / or information. For example, the monitoring device 150 can be used to collect images, videos, sounds, and the like. In some embodiments, the monitoring device 150 can include a data camera, a camera, and the like. In some embodiments, the monitoring device 150 can send the collected data and / or information to the processing device 110 via the network 120, or to the storage device 140 via the network 120.

[0029] The UAV 160 can be used to further monitor the target area. In some embodiments, the UAV 160 can include an unmanned fixed-wing aircraft, an unmanned vertical take-off and landing aircraft, an unmanned airship, an unmanned helicopter, an unmanned multicopter, an unmanned parafoil, etc. In some embodiments, the processing device 110 and / or the terminal 130 can further control the UAV 160 to further monitor the target area when the risk index of the target area in the monitoring image captured by the monitoring device 150 is greater than the first threshold. The UAV monitoring can further determine the public security situation of the target area, avoid misjudgment to cause waste of public security management resources, and improve the efficiency of public security management.

[0030] It should be noted that the application scenarios are provided only for illustrative purposes and are not intended to limit the scope of the present specification. Various modifications or changes can be made according to the description of the present specification for those of ordinary skill in the art. For example, the application scenarios can also include a database. For another example, the application scenarios can be implemented on other devices to implement similar or different functions. However, the changes and modifications will not depart from the scope of the present specification.

[0031] Figure 2 is an exemplary block diagram of a regional public security management early warning system according to some embodiments of the present specification. As shown in Figure 2 The regional public security management early warning system 200 includes a user platform, a service platform, a public security management platform, a sensing network platform, and an object platform. In some embodiments, the smart city regional public security management early warning system 200 can be part of or implemented by the processing device 110.

[0032] In some embodiments, the regional public security management early warning system 200 can be implemented based on an Internet of Things system. In some embodiments, the processing of information in the Internet of Things system can be divided into a processing flow of sensing information and a processing flow of control information, and the control information can be information generated based on the sensing information. Among them, the processing of sensing information is to obtain sensing information by the user platform and deliver it to the public security management platform. The control information is issued by the public security management platform to the user platform, and then the corresponding control is implemented. In some embodiments, when the Internet of Things system is applied to city management, it can be called a smart city Internet of Things system.

[0033] The user platform can be a platform for interacting with users. In some embodiments, the user platform can be configured as a terminal device (e.g., the terminal 130), for example, the terminal device can include a mobile device, a tablet computer, etc., or any combination thereof. In some embodiments, the user platform can be used to receive user input requests and / or instructions. For example, the user platform can obtain a user query request for the risk index of the target area through the terminal device.

[0034] The service platform can be a platform for receiving and transmitting data and / or information. For example, the service platform can send a query request generated by the user platform to the public security management platform. For another example, the service platform can send early warning information generated by the public security management platform to the user platform.

[0035] The public security management platform can refer to a platform that coordinates the contact and cooperation between various functional platforms, aggregates all information of the Internet of Things, and provides sensing management and control management functions for the operation system of the Internet of Things. For example, the public security management platform can obtain the public security situation of a target area (e.g., the presence of suspicious personnel) through the sensing network platform and the object platform, and based on the public security situation, send an instruction to the unmanned aerial vehicle through the public security management platform to further monitor the target area. In some embodiments, the public security management platform can include the processing device 110 and other components in the public security management platform. Figure 1 In some embodiments, the public security management platform can be a remote platform operated by a management personnel, artificial intelligence, or preset rules.

[0036] In some embodiments, the public security management platform can adopt a front-split arrangement. The front-split arrangement can refer to that the public security management platform is provided with a general database and a plurality of management sub-platforms, the plurality of management sub-platforms store, process, and / or transmit corresponding data according to different data sources, each management sub-platform can further aggregate the processed data to the general database, the public security management platform analyzes and processes the aggregated data and stores them, and then transmits the data to the service platform through the general database. In some embodiments, the plurality of management sub-platforms included in the public security management platform can be determined according to preset regions in the city. For example, the public security management platform can include an A-region corresponding management sub-platform, a B-region corresponding management sub-platform, a C-region corresponding management sub-platform, and a plurality of management sub-platforms.

[0037] In some embodiments, in response to the demand of the user for querying the risk index of each region in the city, the public security management platform obtains the monitoring image and other information of the corresponding region from the corresponding sensing network sub-platform of the sensing network platform through the corresponding management sub-platform of the public security management platform, and stores, analyzes, and processes them, thereby determining the risk index of each corresponding region. For example, through the A-region corresponding management sub-platform, the B-region corresponding management sub-platform, and the C-region corresponding management sub-platform of the public security management platform, the risk indexes of the A-region, the B-region, and the C-region are determined respectively and aggregated to the general database of the public security management platform, and the risk indexes of each region are uploaded to the service platform by the general database of the public security management platform, and then uploaded to the user platform by the service platform to feedback to the user. The user can plan and coordinate subsequent arrangements based on the different risk situations of each region.

[0038] The sensing network platform can be a functional platform for managing sensing communication. In some embodiments, the sensing network platform can connect the public security management platform and the object platform, and implement the functions of sensing communication and control information sensing communication. In some embodiments, the sensing network platform can include a plurality of sensing network sub-platforms.

[0039] In some embodiments, the sensing network platform can adopt a standalone arrangement. The standalone arrangement can refer to the sensing network platform adopting different sub-platforms for different types or different data sources of data for data storage, data processing, and / or data transmission. In some embodiments, the plurality of sensing network sub-platforms included in the sensing network platform can be determined according to predetermined areas in the city, which can correspond to the management sub-platforms of the public security management platform. For example, the sensing network platform can set a sensing network sub-platform corresponding to area A, a sensing network sub-platform corresponding to area B, and a sensing network sub-platform corresponding to area C, which respectively correspond to the management sub-platform corresponding to area A, the management sub-platform corresponding to area B, and the management sub-platform corresponding to area C.

[0040] In some embodiments, the sensing network platform responds to the query instructions issued by the management sub-platform of the public security management platform, and obtains monitoring images and other information from the corresponding monitoring devices of the object platform through the corresponding sensing network sub-platform, and uploads them to the corresponding management sub-platform of the public security management platform.

[0041] The object platform can be a functional platform for generating sensing information. In some embodiments, the object platform can be configured to include at least one monitoring device. In some embodiments, the object platform can be used to obtain information related to public security in a target area. For example, based on the monitoring device 150 obtaining monitoring images and other information of different target areas. In some embodiments, the at least one monitoring device configured by the object platform can be provided with a unique identifier (such as a number, etc.) according to a predetermined rule, and the identifier can have a corresponding relationship with the predetermined area, so that the sensing network sub-platform of the sensing network platform can obtain monitoring images and other information of the corresponding area from the corresponding monitoring device. For example, the sensing network sub-platform corresponding to area A can obtain monitoring images of area A from the monitoring device set in area A.

[0042] It is to be noted that the above description of the system and its components is for the convenience of description only and does not limit the present specification to the scope of the embodiments. It can be understood that, after understanding the principles of the system, those skilled in the art can combine the components in any way or connect the components to form a subsystem without departing from the principles. For example, the public security city vaccine management platform and the service platform can be integrated into one component. For another example, the components can share one storage device, and the components can also have their own storage devices. Such variations are within the scope of the present specification.

[0043] Figure 3 is an exemplary flowchart of the Internet of Things-based smart city regional public security management early warning method according to some embodiments of the present specification. In some embodiments, the Internet of Things-based smart city regional public security management early warning method can be performed by the regional public security management early warning system 100 (e.g., the processing device 110) or the regional public security management early warning system 200 (e.g., the public security management platform). For example, the flow 300 can be stored in the form of a program or an instruction in the storage device (e.g., the storage device 140), and when the processing device 110 or the public security management platform executes the program or the instruction, the flow 300 can be implemented. The operation schematic diagram of the flow 300 presented below is illustrative. In some embodiments, the process can be completed with one or more additional operations not described and / or one or more operations not discussed. In addition, Figure 3 The order of the operations of the flow 300 shown in and described below is not limiting.

[0044] At step 310, the user platform obtains the query instruction of the user for each region based on the user platform, and sends the query instruction to the public security management platform via the service platform.

[0045] The query instruction can refer to an instruction request for querying the risk index. In some embodiments, the user can obtain the query instruction of the user for the risk index of each region through the user platform.

[0046] At step 320, in response to the query request, the public security management platform obtains the monitoring image of at least one target region from at least one monitoring device of at least one target region based on the sensing network sub-platform of the sensing network platform.

[0047] The target region can refer to one or more public regions that need to be monitored. For example, indoor or outdoor public regions such as parks, shopping malls, communities, banks, etc. In some embodiments, the target region can include at least one monitoring device deployed in advance.

[0048] The monitoring device can refer to a device for monitoring personnel or behavior in a target area. The monitoring device can be various data collection devices, such as a camera, a video recorder, an image sensor, etc.

[0049] In some embodiments, at least one monitoring device can be configured in the object platform. For example, the monitoring device can be one or more cameras in the object platform.

[0050] The monitoring image can refer to an image or video data in a target area collected by a monitoring device. For example, the monitoring image can be an image, a video, or one or more images in a certain video.

[0051] In some embodiments, the monitoring device can upload the monitoring image to the public security management platform through the sensor network platform for analysis, storage and processing of the monitoring image.

[0052] In some embodiments, the public security management platform and the sensor network platform can be provided with multiple sub-platforms. The multiple sub-platforms can be determined according to multiple predetermined areas in a city. The monitoring image obtained by the monitoring device in the target area can be uploaded to the corresponding management sub-platform of the public security management platform through the corresponding sub-platform of the sensor network platform. For more details, see Figure 2 and the description thereof.

[0053] Step 330, based on the sensor network sub-platform, the monitoring image corresponding to the target area is sent to the management sub-platform.

[0054] In some embodiments, the sensor network sub-platform can obtain monitoring image information from the monitoring device configured in different object platforms and upload it to the management sub-platform of the corresponding public security management platform. For example, the sensor network sub-platform corresponding to area A can obtain the monitoring image of area A from the monitoring device set in area A.

[0055] Step 340, based on the management sub-platform, the monitoring image is processed to determine the risk index of at least one target area.

[0056] The risk index can refer to the possibility of occurrence of various public security events in a target area. For example, the risk index can refer to the possibility of occurrence of personnel fighting, robbery, theft, etc. in a target area. The risk index can be expressed as a value in a predetermined value range, such as a value in the interval [0, 10], or as a predetermined level, such as level 1, level 2, level 3, or various forms such as low, medium, severe.

[0057] In some embodiments, the risk index can be determined based on preset rules. For example, an area with more people is more likely to have a theft incident than an area with fewer people, and the corresponding risk index is higher. For another example, when there is no one in the target area, the risk index can be set to 0, etc. In some embodiments, the risk index can also be determined based on historical public security events in the area, for example, if a certain type of public security event has occurred in the target area for more than 3 years, the risk index of the target area can be considered to be high.

[0058] In some embodiments, the public security management platform can determine a suspicious index of at least one person in the monitoring image, and determine a suspicious person in response to the suspicious index satisfying a preset condition. Further, the public security management platform can determine the risk index of the target area based on the suspicious person. For more information about determining the risk index based on the suspicious person, see Figure 4 and the description thereof.

[0059] In some embodiments, the public security management platform can analyze and process monitoring images of multiple adjacent frames to determine the distance between suspicious persons in the monitoring images. Further, the public security management platform can determine a suspicious group corresponding to the suspicious persons based on the distance, and determine the risk index of the target area based on the suspicious group. For more information about determining the risk index based on the suspicious group, see Figure 6 and the description thereof.

[0060] Step 350, in response to the risk index of the target area being greater than a first threshold, generating an early warning information, and sending the early warning information to the user platform via the general database of the public security management platform and the service platform.

[0061] The first threshold can refer to a threshold corresponding to a risk index preset for determining whether to issue a warning. For example, for a risk index value in the interval [0, 10], the first threshold can be set to 5. The first threshold can be determined based on social experience. For example, it is determined according to the historical public security event situation of the target area. Different target areas can correspond to different first thresholds. For example, a prosperous area with relatively developed economy is more likely to have a social security risk event, and the first threshold can be set to be lower to prevent in advance.

[0062] The early warning information can refer to information for reminding that there may be risks in the target area. The early warning information can be any combination of one or more of text information, sound information, image information, etc. For example, the early warning information can be text information displayed on a terminal device, such as “Please be aware that there are many highly suspicious persons in the mall” and the like; or sound information or audio information played by a warning device or a broadcasting device, etc.

[0063] In some embodiments, each management sub-platform can generate corresponding early warning information for the risk index of each target area, and the total database of the public security management platform can perform summary analysis on the data processed by each management sub-platform. In some embodiments, the summarized data can be sent to the user platform via the service platform, and the user can issue management instructions according to the summarized data at the user platform to perform relevant management (such as whether to send a drone for further monitoring, etc.). The management instruction can refer to an instruction issued by the user for further monitoring of the target area.

[0064] In some embodiments, the public security management platform can also determine a management scheme according to the results of the summary analysis, and send the management scheme to the user platform via the service platform for the user to determine. For example, the management scheme can include whether each target area needs to send a drone for further monitoring, the number of drones to be sent, the frequency of sending drones, etc. Accordingly, the user can review the management scheme at the user platform, and issue corresponding management instructions according to the management scheme that passes the review to perform relevant management. In some embodiments, the public security management platform can send early warning information to the user platform via the service platform to feed back to the user. For example, the public security management platform can feed back the text information of the early warning to the terminal device of the user platform and display it to the user. At step 360, the user platform obtains the management instruction based on the user platform, and controls the drone to go to the target area to monitor the target area according to the management instruction.

[0065] In some embodiments, when the risk index of the target area is greater than the first threshold, the public security management platform can further control the drone to go to the target area to monitor the target area. The drone is installed with a monitoring device (such as a camera, etc.), which can further monitor the target area. For more information about the drone, see Figure 1 and the description.

[0066] In some embodiments, the drone can be controlled by the terminal device (for example, terminal 130). In response to the user accepting the early warning information through the user platform, the user can issue a control instruction to control the drone. For example, the user inputs the control instruction through the terminal device of the user platform. The control instruction can include the target area, the navigation route, etc. The drone can go to the target area and further monitor the target area by executing the control instruction. The drone can track and monitor suspicious persons in the target area under human control; at the same time, the drone can replace the police to go on duty, and the police can perform supervision of the target area in the background of the drone without going on duty.

[0067] In some embodiments of the present specification, the risk index of the target area can be determined in real time based on the monitoring image, so as to improve the timeliness of the early warning; when the risk index exceeds a certain threshold, the early warning can be generated, and the unmanned aerial vehicle can be controlled to go to the corresponding target area for further monitoring, so as to manage the public security of the area and improve the management efficiency.

[0068] In some embodiments, the public security management platform can further acquire a plurality of risk indexes of the target area in a plurality of time periods.

[0069] The plurality of time periods can refer to a plurality of historical time periods up to the current time. For example, the plurality of time periods can be each day in the past week. For another example, the plurality of time periods can be the morning (6:00-11:00), the afternoon (14:00-18:00), the evening (18:00-24:00), etc. in the past three days.

[0070] In some embodiments, the public security management platform can further determine the target area as a daily patrol point in response to the average value of the plurality of risk indexes being greater than a second threshold value.

[0071] The second threshold value can refer to a threshold value of the risk index for determining the daily patrol point. For example, the second threshold value can be 6 or 4, etc. The second threshold value can be set based on a preset rule.

[0072] The daily patrol point can refer to a target area for which the public security personnel are arranged to patrol according to a certain patrol frequency.

[0073] In some embodiments, the public security management platform can determine the daily patrol point based on the situation of the historical public security events of at least one target area. For example, the public security management platform can count the number or frequency of the public security events of a plurality of target areas in a period of time (such as the past month, etc.), and the target area with a higher number or frequency can be set as the daily patrol point.

[0074] In some embodiments, the public security management platform can further determine a plurality of risk indexes of at least one target area in a plurality of time periods based on the monitoring image of the target area, and then calculate the average value of the plurality of risk indexes. When the average value is greater than a second threshold value, the public security management platform can determine that the target area is a daily patrol point.

[0075] In some embodiments, the public security management platform can further determine the patrol frequency of the daily patrol point.

[0076] The patrol frequency can refer to the number of patrols in a preset time period. For example, the patrol frequency can be 5 times a day, 1 time every 2 hours, etc. For another example, the patrol frequency can be 3 times from 06:00 to 11:00, 5 times from 20:00 to 23:00, etc.

[0077] In some embodiments, the patrol frequency can be related to the proportion of suspicious groups with high suspicious indexes in the suspicious group distribution vector. The suspicious groups with high suspicious indexes can be suspicious groups with suspicious indexes greater than a preset threshold. For example, the greater the proportion of suspicious groups with high suspicious indexes in the suspicious group distribution vector, the higher the corresponding patrol frequency. For more information about suspicious groups with high suspicious indexes, see Figure 6 and the description thereof.

[0078] Some embodiments of the present specification can determine the risk index of the target area in multiple dimensions and comprehensively through processing of multiple risk indexes of the target area. When the average value of the risk index is greater than the second threshold, the target area is determined as a daily patrol point, and the monitoring intensity of the target area is increased, which helps to improve the security of the area. In addition, by associating the patrol frequency with the proportion of suspicious groups with high suspicious indexes in the suspicious group distribution vector, the target area can be patrolled more effectively and efficiently, which helps to rationally utilize patrol resources.

[0079] Figure 4 is an exemplary flowchart for determining a risk index based on suspicious personnel according to some embodiments of the present specification. In some embodiments, the method for determining a risk index based on suspicious personnel can be performed by the regional public security management early warning system 100 (e.g., the processing device 110) or the regional public security management early warning system 200 (e.g., the public security management platform). For example, the flow 400 can be stored in the form of a program or instructions in a storage device (e.g., the storage device 140), and when the processing device 110 or the public security management platform executes the program or instructions, the flow 400 can be implemented. The operation schematic of the flow 400 presented below is illustrative. In some embodiments, the process can be completed with one or more additional operations not described and / or one or more operations not discussed. In addition, Figure 4 The order of the operations of the flow 400 shown in and described below is not limiting.

[0080] At step 410, a suspicious index of at least one person in the monitored image is determined.

[0081] The suspicious index can refer to the possibility that the person can cause a public security event. The suspicious index can be a value in the interval [0, 10], such as 0, 5, 8, etc.

[0082] In some embodiments, the public security management platform can determine, through the monitoring image of the target area, the number of times or frequency that the at least one person appears in the target area within a time period, and determine the suspicious index of the person in combination with the public security events or historical statistical data of the target area within the time period. For example, the person appears in the target area every day in the past half year, but no public security event occurs, and the suspicious index of the person is low. For another example, it is identified that the person A appears in the target area, and there is no face information of the person in the historical statistical data, and the suspicious index of the person A can be set to be high, indicating that the person A needs to be observed first. It should be noted that this is not limited to the more times the person appears in the target area, the lower the suspicious index, and the less times the person appears in the target area, the higher the suspicious index.

[0083] In some embodiments, the suspicious index can be related to the area type of the area to which the monitoring device belongs.

[0084] The area type can refer to the attribute or property of the target area. For example, the area type can include bank, jewelry store, supermarket, hotel, etc. For another example, the area type can include square, community, park, commercial street, etc.

[0085] The types, number and frequency of public security events in different types of areas are different. In some embodiments, the different areas to which the monitoring devices belong correspond to different suspicious indexes of the at least one person in the monitoring images taken by the monitoring devices. For example, the monitoring device in the hair salon area takes a picture of person A, and there is no face information of person A in the historical statistical data, and it can be determined that the suspicious index of person A is 5; the monitoring device in the bank area takes a picture of person B, and there is no face information of person B in the historical statistical data, and it can be determined that the suspicious index of person B is 8.

[0086] In some embodiments, the suspicious index of the suspicious person can increase with the increase of the stay duration of the person in the target area.

[0087] The stay duration can refer to the length of time from when the person appears in the target area to when the person leaves the target area. For example, 5 minutes, 60 minutes. The stay duration can also be the cumulative stay duration. For example, after the person leaves the target area and returns to the target area again within a preset time threshold (such as 10 minutes, 5 minutes, etc.), the stay duration of the person can be the sum of the two stay durations.

[0088] In some embodiments, the public security management platform can determine the stay duration of the person according to the image sequence of the multiple monitoring images. For example, the public security management platform can determine multiple monitoring images including a certain person, and determine the stay duration of the person in the target area based on the time difference between the first and last images of the continuous multiple images.

[0089] The suspicious index of the suspicious person is related to the length of stay of the person in the target area, and the relationship can be set based on preset rules. In some embodiments, the relationship curve of the suspicious index and the length of stay can be set according to the experience of public security management, and the public security management platform can determine the suspicious index of the person based on the preset relationship curve by acquiring the length of stay of the person. For example, the suspicious index increases with the increase of the length of stay, reaches a peak when the length of stay increases to a certain degree, and remains unchanged, and then decreases with the increase of the length of stay. It can be understood that the suspicious index of the person changes with the length of stay, and there are growth stage, rapid growth stage, and decline stage, and finally tends to a lower level. It can be understood that the monitoring of the suspicious person will first attract the attention of the public security management personnel, and when the length of stay of the suspicious person reaches a certain degree, the person will be observed by the public security management personnel, and the risk suspicion of causing public security incidents can be gradually eliminated during the period of combination with the patrol situation, that is, the suspicious index is reduced. In some embodiments, the relationship between the suspicious index and the length of stay can also be determined in other various ways, which is not limited in the present specification.

[0090] In some embodiments of the present specification, by considering the correlation between different area types of the target area and the risk index of the target area, the public security management platform can make the evaluation of the risk index of the target area more targeted. In addition, by acquiring the length of stay of the suspicious person, the public security management platform can make the evaluation of the suspicious index of the person more accurate.

[0091] In some embodiments, as the length of stay of the person in the area increases, the suspicious index can increase at a certain growth rate, wherein the growth rate can be related to the area type of the area where the suspicious person stays. For example, when the suspicious person stays near a hair salon, the growth rate of the suspicious index can be 1 / minute, when the suspicious person stays near a bank, the growth rate of the suspicious index can be 2 / minute, and so on.

[0092] In some embodiments, the growth rate of the suspicious index can also be related to the trajectory suspicious degree. For example, within a preset time period, the person repeatedly appears in the target area, that is, the trajectory suspicious degree of the person is high. Correspondingly, the higher the trajectory suspicious degree, the greater the growth rate of the suspicious index of the person.

[0093] The trajectory suspicious degree can refer to the degree of suspiciousness of the action trajectory of the person. The trajectory suspicious degree can be a value in the interval [0, 1], such as 0.5. The action trajectory of the person includes information such as the appearance of the person in different areas, the length of stay in different areas, and the number of times.

[0094] In some embodiments, the public security management platform can obtain an action track of each person in the monitoring image, and extract a track feature based on the action track. Then, the public security management platform can determine a track suspicious degree based on the track feature. For more details about determining the track suspicious degree, see Figure 5 and the description thereof.

[0095] In step 420, a suspicious person is determined in response to the suspicious index satisfying a preset condition.

[0096] The preset condition can refer to a preset condition for determining whether a person is a suspicious person. For example, the suspicious index is a value in the interval [0, 10], and the preset condition can be 7.

[0097] In some embodiments, the public security management platform can compare and analyze the suspicious index of at least one person with a preset condition. Then, in response to the suspicious index satisfying the preset condition, the public security management platform can determine that the person is a suspicious person. Wherein, satisfying the preset condition can refer to that the suspicious index is greater than or equal to the preset condition.

[0098] In step 430, a risk index of a target area is determined based on the suspicious person.

[0099] In some embodiments, the public security management platform can determine the risk index of the target area based on the number of suspicious persons in the target area. For example, the more the number of suspicious persons, the higher the risk index of the target area can be determined. For example only, the risk index of the target area can be the sum of the suspicious indexes of all suspicious persons.

[0100] In some embodiments of the present specification, the risk index of the target area is determined by comprehensively considering the risk index of each of the plurality of persons in the monitoring image, which helps to accurately evaluate the risk index of the target area.

[0101] Figure 5 is an exemplary schematic diagram of determining a track suspicious degree according to some embodiments of the present specification.

[0102] In some embodiments, as Figure 5 shown, the public security management platform can obtain an action track 520 of each person in the monitoring image 510.

[0103] The action track can refer to a path of staying and moving between a plurality of locations within a certain period of time. For example, the action track of a person can be "location A→location B→location C→location D→location C→location B", which means that the person passes through location A, location B, location C, location D in turn, and then returns to location C from location D, and then returns to location B from location C. Wherein, the plurality of locations can be locations within the same target area, or can be locations within different target areas.

[0104] In some embodiments, the action trajectory can be obtained based on monitoring devices at multiple locations in the target area. The public security management platform can perform face recognition or face matching on the monitoring images obtained at the multiple locations to identify the personnel in each frame of the monitoring images, so as to determine whether the corresponding personnel appears in different locations. Meanwhile, the action trajectory of the personnel can be determined according to the order in which the personnel appears at different locations, the time when the personnel appears, the time when the personnel leaves, and the like.

[0105] In some embodiments, as shown in FIG. 5B, the public security management platform can extract trajectory features based on the action trajectory 520. Figure 5

[0106] The trajectory features can refer to features of the action trajectory of the personnel. The trajectory features can include the target areas visited by the personnel, the number of times of visiting each target area, the length of stay, and the like. It should be noted that the trajectory features can also include time sequence features. For example, the time sequence of the personnel to visit each target area.

[0107] In some embodiments, the public security management platform can determine the trajectory features by modeling or using various data analysis algorithms. For example only, the public security management platform can determine the trajectory features such as the number of times of appearing at different locations and the length of stay by a statistical method. For example, the personnel m appears at location A at 16:00, leaves location A at 16:08, then appears at location B at 16:10, leaves location B at 16:30, and finally appears at location A again at 16:35 and leaves location A at 17:00. It can be seen that the personnel m appears at location A twice, with the length of stay being 8 minutes and 25 minutes respectively, and appears at location B once, with the length of stay being 20 minutes.

[0108] In some embodiments, the trajectory features can be embodied by a trajectory graph. The trajectory graph can be a data structure composed of nodes and edges, and the edges connect the nodes. The nodes and the edges can have attributes.

[0109] In some embodiments, the public security management platform can construct a trajectory graph based on the action trajectory of the personnel.

[0110] The nodes of the trajectory graph can correspond to each location in the action trajectory of the personnel. For example, the nodes can represent locations such as parks, banks, and squares. The node attributes of the trajectory graph can include the length of stay of the personnel at the location, and the like. In some embodiments, the node attributes of the trajectory graph can also include the time when the personnel appears at and leaves the corresponding location. For example, the node attributes of node A can be (16:02, 16:18, 16), indicating that the personnel appears at node A at 16:02 and leaves at 16:18, with the length of stay being 16 minutes. If the personnel appears at a location multiple times, there are multiple times of appearing, multiple times of leaving, and multiple lengths of stay. ​

[0111] An edge of the trajectory graph can refer to the relationship between two connected target areas. The edge of the trajectory graph can be a one-way edge, and the direction of the edge represents the movement of the person from one place to another. The attribute of the edge can include the number of times the person moves from one place to another within a period of time.

[0112] As shown in FIG. 5B, the nodes in the trajectory graph 530 include place A, place B, place C, and place D. The edge AB can be generated from the direction of place A to place B, and the attribute of the edge AB is 1, indicating that the person moves from place A to place B once within a period of time. The edge BC can be generated from the direction of place B to place C, and the attribute of the edge BC is 3, indicating that the person moves from place B to place C three times within a period of time. In addition, the edge CB can be generated from the direction of place C to place B, and the attribute of the edge CB is 2, indicating that the person moves from place C to place B three times within a period of time. Figure 5 In some embodiments, the public security management platform can determine the action trajectory of the person based on the attributes of the nodes and the directions of the edges. For example, the node attributes of node A are (16:02, 16:18, 16), (16:30, 16:42, 12), and the node attributes of node B are (16:20, 16:28, 8), and it can be determined that the action trajectory of the person is “node A→node B→node A”.

[0113] In some embodiments, the public security management platform can determine the trajectory suspiciousness 550 based on the trajectory graph 530. Specifically, the public security management platform can process the trajectory graph 530 through the trajectory suspiciousness determination model 540 to determine the trajectory suspiciousness 550.

[0114] The trajectory suspiciousness can refer to the degree of suspiciousness of the action trajectory of the person. It can be understood that the trajectory suspiciousness of the person can represent the probability that the person may cause a public security event. The trajectory suspiciousness can be a value in the interval [0, 1], and the larger the value, the higher the trajectory suspiciousness. It can also be represented in the form of grades, such as normal, slight, moderate, severe, etc.

[0115] The trajectory suspiciousness determination model can refer to a model for determining the suspiciousness of the action trajectory. In some embodiments, the trajectory suspiciousness determination model can be a graph neural network model (GNN). As shown in FIG. 5C, the input of the trajectory suspiciousness determination model 540 can be the trajectory graph 530, and the output can be the trajectory suspiciousness 550.

[0116] Figure 5 The trajectory suspiciousness determination model can also be other graph models, such as a graph convolutional neural network model (GCNN), or adding other processing layers to the graph neural network model or modifying the processing method thereof.

[0117] The trajectory suspiciousness determination model can also be other graph models, such as a graph convolutional neural network model (GCNN), or adding other processing layers to the graph neural network model or modifying the processing method thereof.​

[0118] In some embodiments, the trajectory suspiciousness determination model 540 can output the trajectory suspiciousness 550 from a current node in the trajectory graph 530. The current node can be a node corresponding to a last stay location of the person. For example, if the node C in the trajectory graph is a node corresponding to a last stay location of the person, the trajectory suspiciousness is output from the node C.

[0119] In some embodiments, the trajectory suspiciousness can be updated based on multiple iterations. For example, the public security management platform can update the action trajectory 520 of the person based on the continuously obtained monitoring image 510, and further update the attributes of the nodes and edges in the trajectory graph 530. Accordingly, the trajectory suspiciousness determination model 540 can be updated once for each update of the trajectory suspiciousness 550.

[0120] In some embodiments, the trajectory suspiciousness determination model can be obtained by training multiple labeled training samples. For example, the multiple labeled training samples can be input to an initial trajectory suspiciousness determination model, a loss function can be constructed based on the labels and the output of the initial trajectory suspiciousness determination model, the parameters of the trajectory suspiciousness determination model can be updated based on the loss function, and the training is completed when a preset condition is met, and a trained trajectory suspiciousness determination model is obtained. The preset condition can be that the loss function is less than a threshold, converges, or the training period reaches a threshold.

[0121] In some embodiments, the training samples can be multiple trajectory graphs of the same structure constructed based on the action trajectories of multiple different persons. The data sources of the action trajectories of the multiple different persons can be obtained based on historical monitoring images of multiple target regions. The labels of the training samples can be the trajectory suspiciousness corresponding to each action trajectory graph. In some embodiments, the labels can be determined by analyzing and judging the corresponding action trajectory graph based on experience or experts. The labels of the training samples can be manually labeled.

[0122] In some embodiments of the present specification, when predicting based on the trajectory suspiciousness determination model, the connection between the nodes and the edges can be further considered, so that the trajectory suspiciousness determination model has a better learning effect on the features, to improve the accuracy of determining the trajectory suspiciousness.

[0123] In some embodiments, the public security management platform can also analyze and process the track features according to preset rules to determine the track suspicious degree. For example, one or more of any combination of preset stay duration, number of appearances in the target area, time interval of appearances, etc. can be used to determine the track suspicious degree. For example, if a person appears near a bank more than 5 times within 24 hours, the corresponding track suspicious degree can be preset as 2 (slightly suspicious); if the average stay duration of the person in the area is 20 minutes, the corresponding track suspicious degree can be preset as 5 (moderately suspicious); if the interval between each appearance of the person is 15 minutes, it indicates that the person is wandering in the area and may be a person who is squatting to withdraw money, and the corresponding track suspicious degree can be preset as 8 (serious).

[0124] Some embodiments of the present specification consider the actual situation of the stay duration of multiple path locations and the track direction in the action track when determining the track suspicious degree of the action track, which can assist in determining whether the person is likely to cause a risk of public security incidents, help the public security management platform accurately and quickly determine the track suspicious degree of the person, and also help further accurately assess the suspicious index of the person.

[0125] Figure 6 is an exemplary flowchart of determining a risk index based on a suspicious group according to some embodiments of the present specification. In some embodiments, the method of determining a risk index based on a suspicious group can be performed by the regional public security management early warning system 100 (e.g., the processing device 110) or the regional public security management early warning system 200 (e.g., the public security management platform). For example, the flow 600 can be stored in the form of a program or instructions in a storage device (e.g., the storage device 140), and when the processing device 110 or the public security management platform executes the program or instructions, the flow 600 can be implemented. The operation schematic diagram of the flow 600 presented below is illustrative. In some embodiments, the process can be completed with one or more additional operations not described and / or one or more operations not discussed. In addition, Figure 6 The order of the operations of the flow 600 shown in FIG. 6 and described below is not limiting.

[0126] At step 610, the distance between suspicious persons is determined based on the monitoring images of multiple adjacent frames.

[0127] The distance can be the straight-line distance between two suspicious persons, which can be determined by the length of the line segment connecting the two. In some embodiments, the public security management platform can determine the distance between suspicious persons in various ways. For example, the distance between suspicious persons can be determined by a distance estimation model, an image recognition algorithm, etc.

[0128] At step 620, the suspicious group corresponding to the suspicious persons is determined based on the distance.

[0129] A suspicious group refers to a plurality of suspicious persons who are gangsters. The plurality of suspicious persons who are gangsters can often appear in one place, have consistent paths, and the like. In some embodiments, there can be a plurality of suspicious groups in the plurality of suspicious persons.

[0130] In some embodiments, the public security management platform can determine whether the two suspicious persons belong to the same suspicious group by the distance between the two suspicious persons in a plurality of adjacent frames. If the distance between the two suspicious persons in the adjacent frames is less than a preset distance threshold (for example, 2 meters), the frame number is further recorded, and if the frame number is greater than a preset frame number, the public security management platform can determine that they belong to the same suspicious group. The preset distance threshold and the preset frame number can be artificially set or set according to the distance of the suspicious persons in the historical suspicious group and the corresponding continuous frame number. For example, the distance threshold is artificially set to 2 meters, the preset frame number is 30 frames, and the distance between suspicious persons A and B in 35 continuous frames is less than 2 meters, then it is determined that suspicious persons A and B belong to the same suspicious group. In some embodiments, the public security management platform can determine whether the two suspicious persons belong to the same suspicious group by the advancing direction between the two suspicious persons in a plurality of adjacent frames. For example, if the advancing direction of suspicious persons A and B in a plurality of continuous frames is the same, it is determined that they belong to the same suspicious group. In some embodiments, whether the two suspicious persons belong to the same suspicious group can be determined by comprehensively considering the distance and the advancing direction between the two suspicious persons in a plurality of adjacent frames. For example, if the advancing direction of suspicious persons A and B in a plurality of continuous frames is the same and the distance is close (for example, walking side by side), it is determined that they belong to the same suspicious group.

[0131] Step 630, determining the risk index of the target area based on the suspicious group.

[0132] In some embodiments, the public security management platform can determine the risk index of the target area based on the number of suspicious groups. The more the number of suspicious groups, the higher the risk index of the target area. Exemplarily, the risk index can be the product of the number of suspicious groups and a preset parameter, wherein the preset parameter can be determined according to the historical number of suspicious groups and the corresponding historical risk index. For example, when the historical number of suspicious groups is in the number interval 1-3, the corresponding historical risk index is 0.3, and when the public security management platform determines that the number of suspicious groups of the target area A is 2, the corresponding preset parameter can be 0.3.

[0133] In some embodiments of the present specification, by setting the preset parameter, the influence of different numbers of suspicious groups on the risk index can be comprehensively considered, and the accuracy of determining the risk index can be improved.

[0134] In some embodiments, the public security management platform can determine the risk index of the target area based on the sum of suspicious group indexes of each suspicious group in the target area and a multiplier factor. The suspicious group index can be the product of the sum of suspicious indexes of each suspicious person in the group and the multiplier factor.

[0135] The multiplier factor can reflect the number of suspicious persons in the suspicious group. For example, the more suspicious persons in the suspicious group, the greater the value of the multiplier factor. In some embodiments, the multiplier factor can be determined according to the number of suspicious persons in the suspicious group. For example, when the number of suspicious persons is 1, the number multiplier factor can be 1, and accordingly, the risk index of the target area can be determined based on the suspicious person. For more information about determining the risk index based on the suspicious person, see Figure 4 and the related description, which will not be repeated here.

[0136] Some embodiments of the present specification help to evaluate the probability of public security incidents in the target area by considering multiple suspicious groups and determining the influence degree of each group according to the number of persons in each suspicious group, thereby more accurately evaluating the risk index of the target area.

[0137] In some embodiments, the risk index of the target area can also be related to the distribution vector of the suspicious groups in the target area.

[0138] The distribution vector of the suspicious group is used to represent the number of suspicious groups and their suspicious indexes. The vector can be a distribution vector designed based on the bucketing principle. For example, a suspicious group distribution vector is (1, 0, 2), which can represent that there is 1 suspicious group with a suspicious index of 0-40, 0 suspicious group with a suspicious index of 40-70, and 2 suspicious groups with a suspicious index of 70-100.

[0139] In some embodiments, the risk index of the target area can also be the product of the sum of suspicious group indexes of each suspicious group and a weight.

[0140] In some embodiments, the weight can be determined based on a preset rule table. It can be understood that 2 suspicious groups with a suspicious index of 50 and 1 suspicious group with a suspicious index of 100 have the same total suspicious index, but the risk index of the area should not be the same because the suspicious group with a suspicious index of 100 leads to a higher probability of public security incidents than the 2 suspicious groups with a suspicious index of 50. Accordingly, the design rule of the preset rule table is that the greater the proportion of suspicious groups with high suspicious indexes in the suspicious group distribution vector, the greater the weight. The suspicious group with a high suspicious index refers to the suspicious group with a suspicious index greater than a preset threshold. For example, the preset threshold is 70, that is, the suspicious group with a suspicious index greater than 70 is the suspicious group with a high suspicious index.

[0141] For example, when the suspicious group distribution vector is (1, 1, 0), it means that there is one suspicious group with a suspicious index of 0-40, one suspicious group with a suspicious index of 40-70, and no suspicious group with a suspicious index of 70-100, i.e., the proportion of suspicious groups with a high suspicious index is 0, and the corresponding weight can be 1.1; when the suspicious group distribution vector is (0, 2, 2), it means that there is no suspicious group with a suspicious index of 0-40, two suspicious groups with a suspicious index of 40-70, and two suspicious groups with a suspicious index of 70-100, i.e., the proportion of suspicious groups with a high suspicious index is 50%, and the corresponding weight can be 1.2.

[0142] In some embodiments of the present specification, by judging whether the suspicious person belongs to a suspicious group, the risk index of the target area can be further determined according to the suspicious group. By considering multiple suspicious groups and determining the influence degree of each group according to the number of people in each suspicious group, the risk index of the target area is evaluated, which helps to evaluate the probability of public security incidents in the target area, and further more accurately evaluate the risk index of the target area. At the same time, by designing the distribution vector of the suspicious group based on the bucketing principle, the risk index can be further determined according to the proportion of suspicious groups with a high suspicious index, so that the risk index of the suspicious group is more accurate.

[0143] It should be noted that the above description of the process 600 is only for example and illustration, and does not limit the scope of the present specification. Those skilled in the art can make various modifications and changes to the process 600 under the guidance of the present specification. However, these modifications and changes are still within the scope of the present specification.

[0144] The above has described the basic concept, and it is obvious that the above detailed disclosure is only for example and does not constitute a limitation on the present specification. Although it is not explicitly stated here, those skilled in the art can make various modifications, improvements and modifications to the present specification. Such modifications, improvements and modifications are suggested in the present specification, so such modifications, improvements and modifications are still within the spirit and scope of the exemplary embodiments of the present specification.

[0145] At the same time, the present specification uses specific words to describe the embodiments of the present specification. As "one embodiment", "an embodiment", and / or "some embodiments" means a certain feature, structure or characteristic related to at least one embodiment of the present specification. Therefore, it should be emphasized and noted that the "an embodiment" or "one embodiment" or "one alternative embodiment" mentioned in different places in the present specification does not necessarily refer to the same embodiment. In addition, some features, structures or characteristics in one or more embodiments of the present specification can be properly combined.

[0146] Furthermore, the order of the processing elements and sequences described in this specification are not intended to be construed as a limitation, unless specifically stated, but are included to provide a complete description of one or more embodiments of the present specification. Regardless of the particular sequence of processing elements and sequences, however, the description herein of a process should be understood to include any and all combinations of one or more elements of a process independently selected from each sequence. For example, although the system components described above can be implemented by hardware devices, they can also be implemented by software solutions, such as installing the described system on an existing server or mobile device.

[0147] Similarly, it is to be noticed that the term "comprising", used in the description, is not intended to exclude other elements or steps. It is to be understood that the description and the examples are intended to be illustrative, but not limiting, of the scope of the present specification. Thus, the scope of the present specification should be given by the appended claims, along with their full scope of equivalents, and not by an restricting interpretation of the description or the examples.

[0148] Some embodiments use numerical designations to describe components, quantities of attributes. It is to be understood that such numerical designations used in the description of embodiments are, in some examples, modified by the adjectives "about", "approximately", or "generally". Unless otherwise stated, "about", "approximately", or "generally" indicates that the stated numerical value is allowed ±20% variation. Accordingly, numerical values used in the description and claims are approximations that can vary depending on the desired properties of the individual embodiments. In some embodiments, numerical values should be considered in the context of the number of significant figures used in the description and claims. Although the numerical ranges and parameters setting forth the broadest scope of the embodiments herein are approximations, the numerical values set forth in the specific examples are reported as precisely as practicable. The numerical values set forth in the specific examples are provided to give a general understanding of the embodiments.

[0149] Each patent, patent application, patent publication, and other material cited in this specification is hereby incorporated by reference in its entirety. In the event of inconsistencies between the disclosure of this specification and the materials incorporated by reference, the disclosure of this specification is intended to prevail. In the event of inconsistencies between the disclosure of this specification and the claims, the claims are intended to prevail. It is specifically noted that, where a description, definition, and / or a term in the materials incorporated by reference in this specification is inconsistent or in conflict with that in the specification, the description, definition, and / or term in the specification is intended to prevail.

[0150] Finally, it should be understood that the embodiments described herein are only given by way of example and that other modifications can occur to persons skilled in the art. Therefore, the scope of the present description is not intended to be limited to the embodiments described herein but is only limited by the claims that follow.

Claims

1. A method for generating risk area early warning information for smart cities, characterized in that, The method is executed by the public security management platform and includes: Retrieve user query commands for each region; Based on the query command, obtain surveillance images of at least one target area; The surveillance image is processed to determine the suspiciousness index of at least one person in the surveillance image. The suspiciousness index is determined based on the area type of the area to which the surveillance device belongs. When the personnel suspiciousness index meets preset conditions, suspicious personnel are identified. Based on the surveillance images of multiple adjacent frames, the distance between the suspicious persons is determined; If the distance between two suspicious persons in adjacent frames is less than a preset distance threshold, the number of frames is recorded. If the number of frames is greater than a preset number of frames, it is determined that the suspicious persons in the multiple adjacent frames belong to the same suspicious group. The suspicious group includes multiple suspicious persons who commit crimes as a gang. If two suspicious persons in the plurality of adjacent frames have the same direction of travel, then it is determined that the suspicious persons in the plurality of adjacent frames belong to the same suspicious group. If the distance between two suspicious persons in the adjacent frames is less than a preset distance threshold and the two suspicious persons in the multiple adjacent frames have the same direction of travel, then it is determined that the suspicious persons in the multiple adjacent frames belong to the same suspicious group. The risk index of the target area is determined based on the sum of the suspicious group indices of each suspicious group in the target area; the suspicious group index is the product of the sum of the personnel suspicious indices of each suspicious person in the suspicious group and the amplification factor, and the amplification factor is positively correlated with the number of suspicious persons in the suspicious group. A warning message is generated in response to the risk index of the target area being greater than a first threshold.

2. The method according to claim 1, characterized in that, The determination of the risk index of the target area based on the sum of the suspicious group indices of each suspicious group in the target area includes: Based on the sum and weight of the suspicious group index for each suspicious group in the target area, the risk index of the target area is determined, wherein... The weights are determined based on a preset rule table, which is based on the distribution vector of suspicious groups.

3. The method according to claim 1, characterized in that, Also includes: Obtain multiple risk indices for the target area over multiple time periods; In response to the average of the multiple risk indices being greater than a second threshold, the target area is determined to be a routine patrol point.

4. The method according to claim 1, characterized in that, The public security management platform includes a main database and multiple management sub-platforms, and the multiple management sub-platforms included in the public security management platform are determined according to the preset areas in the city; The query command is obtained through the user platform and sent to the public security management platform via the service platform. The surveillance images of the at least one target area are acquired by the security management platform from at least one monitoring device in the at least one target area through a sensor network sub-platform based on the sensor network platform; the at least one monitoring device is configured in different object platforms; wherein, the sensor network platform uses different sensor network sub-platforms for data storage, data processing and / or data transmission for data from different object platforms, and the sensor network sub-platforms correspond to different target areas; the security management platform uses different management sub-platforms for data storage, data processing and / or data transmission, and performs data aggregation, data processing and data transmission through the overall database of the security management platform; The monitoring image of the at least one target area is sent from the sensor network sub-platform to the management sub-platform; The early warning information is sent to the user platform via the overall database of the public security management platform and the service platform.

5. The method according to claim 1, characterized in that, The personnel suspiciousness index is positively correlated with the duration of stay of the at least one person in the target area; the growth rate of the personnel suspiciousness index is related to the type of area where the at least one person stayed and the suspiciousness of their trajectory; wherein, the suspiciousness of the trajectory is determined in the following manner: Obtain the movement trajectory of at least one person in the surveillance image; Based on the movement trajectory, trajectory features are extracted. These trajectory features are represented by a trajectory graph. The nodes of the trajectory graph correspond to various locations. The attributes of the nodes include the time points when the person appears and leaves. The edges of the trajectory graph are unidirectional edges. The direction of the unidirectional edge represents the person moving from one location to another. The attributes of the edge include the number of times the person has moved. The trajectory map is processed based on the trajectory suspicion determination model to determine the trajectory suspicion, wherein the trajectory suspicion determination model is a graph neural network model, and the trajectory suspicion is based on the node output corresponding to the last location where the person stayed; In response to the personnel suspiciousness index meeting preset conditions, the suspicious personnel are identified; Based on the suspected individuals, the risk index of the target area is determined.

6. An Internet of Things (IoT) system for generating risk area early warning information for smart cities, characterized in that, The Internet of Things (IoT) system includes a user platform, a service platform, a security management platform, a sensor network platform, and an object platform that interact sequentially. The security management platform is configured to perform the following operations: Based on the user platform, the query instructions of the user for each region are obtained, and the user platform is configured as at least one terminal device. In response to the query command, the security management platform acquires monitoring images of at least one target area from at least one monitoring device in at least one target area based on the sensor network sub-platform of the sensor network platform. The at least one monitoring device is configured in different target platforms. The sensor network platform uses different sensor network sub-platforms for data storage, processing, and / or transmission for data from different target platforms, and each sensor network sub-platform corresponds to a different target area. The security management platform uses different management sub-platforms for data storage, processing, and / or transmission, and aggregates, processes, and transmits data through the overall platform of the security management platform. Based on the sensor network sub-platform, the monitoring image of the corresponding target area is sent to the management sub-platform; The monitoring images are processed based on the management sub-platform to determine the suspiciousness index of at least one person in the monitoring images; the suspiciousness index is determined based on the area type of the area to which the monitoring equipment belongs. When the personnel suspiciousness index meets preset conditions, suspicious personnel are identified. Based on the surveillance images of multiple adjacent frames, the distance between the suspicious persons is determined; If the distance between two suspicious persons in adjacent frames is less than a preset distance threshold, the number of frames is recorded. If the number of frames is greater than a preset number of frames, it is determined that the suspicious persons in the multiple adjacent frames belong to the same suspicious group. The suspicious group includes multiple suspicious persons who commit crimes as a gang. If two suspicious persons in the plurality of adjacent frames have the same direction of travel, then it is determined that the suspicious persons in the plurality of adjacent frames belong to the same suspicious group. If the distance between two suspicious persons in the adjacent frames is less than a preset distance threshold and the two suspicious persons in the multiple adjacent frames have the same direction of travel, then it is determined that the suspicious persons in the multiple adjacent frames belong to the same suspicious group. The risk index of the target area is determined based on the sum of the suspicious group indices of each suspicious group in the target area; the suspicious group index is the product of the sum of the personnel suspicious indices of each suspicious person in the suspicious group and the amplification factor, and the amplification factor is positively correlated with the number of suspicious persons in the suspicious group. A warning message is generated in response to the risk index of the target area being greater than a first threshold.

7. A device for generating risk area early warning information for smart cities, characterized in that, The risk area early warning information generation device for smart cities includes a processor, which is used to execute the method of any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The storage medium stores computer instructions, and when the computer reads the computer instructions, the computer executes the method as described in any one of claims 1 to 5.

Images