A method for crowd alarm based on personnel positioning scenarios

Abstract

The application discloses a kind of based on personnel positioning scene's gathering alarm method, the present application relates to gathering alarm technical field, solve the personnel positioning system usually needs to rely on specific hardware equipment, this not only increases the cost of system, also limit its application flexibility and popularization rate The problem of problem, the present application can not only judge the gathering behavior of each area by the method, but also send the relevant picture to display module when detecting gathering behavior, this real-time alarm mechanism allows staff to understand gathering situation in initial stage, timely take measures to prevent potential safety risk, in emergency, fast and accurate response has vital role in avoiding situation to expand, simultaneously, current gathering situation can be shown by visual way, to help staff more intuitively understand on-site condition, improve the efficiency and accuracy of decision-making.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of crowd alarm, in particular to a crowd alarm method based on personnel positioning scenarios. BACKGROUND

[0002] In personnel positioning scenarios, existing technologies mainly rely on various sensors and positioning systems such as GPS, Wi-Fi, Bluetooth, RFID, and other wireless communication technologies to track and determine the location of individuals. These technologies can monitor the location changes of personnel in real time and are widely used in security monitoring, elderly and child care, employee management, and other fields. For example, in large public events or building sites, these positioning technologies can be used to monitor personnel distribution in real time, effectively manage pedestrian flow, and prevent potential safety risks.

[0003] The application with publication number CN116704723A discloses a personnel gathering detection method, which has the following advantages: by equipping personnel with a positioning device to generate a spherical space, the number of surrounding personnel is obtained, and the personnel density is obtained, thereby effectively controlling the degree of personnel gathering in any area; by adjusting the detection radius of each detection point according to the danger level of the personnel's work, and because of the error of the positioning device, the alarm range is divided into a core alarm area and a non-core alarm area, improving the safety of detection; according to the intersection of different alarm areas caused by personnel flow, the alarm situation is classified, and corresponding emergency measures are taken according to the alarm level, thereby reducing personnel casualties and other consequences caused by safety accidents.

[0004] The existing technology still has some deficiencies in practical application. Personnel positioning systems usually need to rely on specific hardware devices, which not only increases the cost of the system, but also limits the flexibility and popularity of its application. Therefore, developing a new crowd alarm method that can accurately and timely judge the crowd gathering situation is of great significance for improving public safety and preventing potential risks. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a crowd alarm method based on personnel positioning scenarios, which solves the problems in the background art.

[0006] To achieve the above purpose, the present application realizes the following technical scheme: a crowd alarm method based on personnel positioning scenarios, comprising:

[0007] A data acquisition module is used to collect the number of people in the target area in real time, obtain the specific area of the target area, and send the number of people to the subsequent time period division module, wherein the target area is the area that can be shot by multiple cameras;

[0008] The period division module is configured to receive the number of people information in real time, divide a day into a plurality of time periods, determine the crowd density of each time period according to the number of people information, then select a time period as a target time period in sequence, obtain the crowd density of the target time period in the past n days from the current time, determine the target time period as a high-frequency time period or a low-frequency time period through a preset formula, and thereby divide the plurality of time periods of a day into high-frequency time periods or low-frequency time periods, wherein n is a preset value.

[0009] The region division module is configured to divide the target region into a plurality of small regions through a preset region division model, determine an analysis time period, sequentially take the plurality of small regions as analysis regions, and determine whether the analysis region is a key region or a non-key region, wherein the analysis time period is determined in the high-frequency time period or the low-frequency time period.

[0010] The region determination module is configured to divide the target region into four types of regions, i.e., a high-frequency time period key region, a high-frequency time period non-key region, a low-frequency time period key region, and a low-frequency time period non-key region, according to the period division module and the region division module.

[0011] The gathering behavior alarm module is configured to determine whether each type of region appears a gathering behavior according to the four types of regions, and when it is determined that the gathering behavior appears, issue a gathering alarm and send a shooting picture to a subsequent display module.

[0012] Preferably, the determination method of the high-frequency time period and the low-frequency time period in the period division module includes the following steps.

[0013] S1: divide a day into y time periods, wherein y is a preset value.

[0014] S2: arbitrarily select a time period as a target time period, divide the target time period into x hour periods on average, obtain the specific number of people Rk in the target region at the end of each hour period according to the number of people information, then obtain the specific area S of the target region, obtain the crowd density Gk of the target region at this time through calculation Gk=Rk / S, and then obtain the result U as the crowd density of the target time period through the formula wherein x is a preset value, 1≦k≦x.

[0015] S3: then obtain the crowd density Ui of the target time period in the past n days from the current time, wherein n>1, 1≦i≦n, and n=90 in this embodiment.

[0016] S4: then calculate the average value Up of the n groups of crowd density Ui, i.e., S5: then calculate the deviation value H of the n groups of crowd density Ui, i.e., S6: determine the size of the deviation value H.

[0017] Preferably, in step S6, it further includes:

[0018] S61: When H≤D1, the average value Up is taken as the crowd density value Usp of the target time period, and the crowd density value Usp is compared with the preset value D2, when Usp>D2, the target time period is determined as a high-frequency time period, and when Usp≦D2, the target time period is determined as a low-frequency time period;

[0019] S62: When H>D1, it is determined that the deviation value H is too large, Ui is sorted in descending order of |Ui-Up|, and Ui sorted first is deleted each time and the deviation value H is recalculated until H≦D1;

[0020] Then, the number D3 of the remaining Ui is counted, and the ratio of D3 to n is calculated and compared with the preset ratio D4:

[0021] When the ratio of D3 to n is greater than the preset ratio D4, the average value of the remaining Ui is calculated as the crowd density value Usp of the target time period, and the crowd density value Usp is compared with the preset value D2, when Usp>D2, the target time period is determined as a high-frequency time period, and when Usp≦D2, the target time period is determined as a low-frequency time period;

[0022] When the ratio of D3 to n is less than or equal to the preset ratio D4, the average value of the deleted Ui is calculated as the crowd density value Usp of the target time period, and the crowd density value Usp is compared with the preset value D2, when Usp>D2, the target time period is determined as a high-frequency time period, and when Usp≦D2, the target time period is determined as a low-frequency time period;

[0023] Wherein, D1 and D2 are preset values.

[0024] Preferably, after step S6, it further comprises:

[0025] S7: Then, steps S2-S6 are repeated to divide y time periods into high-frequency time periods or low-frequency time periods respectively.

[0026] Preferably, in the area division module, the division of the key area and the non-key area comprises the following steps:

[0027] P1: The target area is evenly divided into m small areas, each small area has the same area, denoted as Sj, 1≦j≦m, and P2: A high-frequency time period or a time period with the largest crowd density value Usp when there is no high-frequency time period is selected as an analysis time period, and the analysis time period is evenly divided into x time periods;

[0028] P3: randomly select one of the m small areas as an analysis area, at the end of each hour segment, obtain the specific number of people Rq in the analysis area, then calculate the crowd density Gq=Rq / Sj of the analysis area at this time, and then obtain the specific crowd density Zj of the analysis area in the analysis period through the formula wherein 1≦q≦x. Preferably, after step P3, it further comprises:

[0029] P4: then repeat P3 to obtain the specific crowd density Zj of all m small areas as analysis areas in the analysis period, and then compare the crowd density Zj with a preset value D5.

[0030] Preferably, in step P4, it further comprises:

[0031] When Zj>D5, the small area is determined as a key area, and when Zj≦D5, the small area is determined as a non-key area, wherein D5 is a preset value, which is specifically set by staff;

[0032] Preferably, in the gathering behavior alarm module, the step of determining whether the gathering behavior occurs in each type of area comprises the following steps:

[0033] AS1: in the key area during the high-frequency period, the specific number of people is collected every T1 time period in real time, when the number of people collected for the first time exceeds L1, it is marked as "1", and when it is marked as "1" for three times in succession, it is determined as a gathering behavior, an alarm of gathering is issued and the shooting picture is sent to the subsequent display module;

[0034] AS2: in the non-key area during the high-frequency period, the specific number of people is collected every T2 time period in real time, when the number of people collected for the first time exceeds L2, it is marked as "1", and when it is marked as "1" for three times in succession, it is determined as a gathering behavior, an alarm of gathering is issued and the shooting picture is sent to the subsequent display module;

[0035] AS3: in the key area during the low-frequency period, the specific number of people is collected every T3 time period in real time, when the number of people collected for the first time exceeds L3, it is marked as "1", and when it is marked as "1" for three times in succession, it is determined as a gathering behavior, an alarm of gathering is issued and the shooting picture is sent to the subsequent display module;

[0036] AS4: in the non-key area during the low-frequency period, the specific number of people is collected every T4 time period in real time, when the number of people collected for the first time exceeds L4, it is marked as "1", and when it is marked as "1" for three times in succession, it is determined as a gathering behavior, an alarm of gathering is issued and the shooting picture is sent to the subsequent display module;

[0037] Wherein, T1, T2, T3 and T4 are preset values, and T4>T3>T2>T1, L1, L2, L3 and L4 are preset values, and L1>L2>L3>L4.

[0038] Preferably, it further comprises:

[0039] The gathering behavior special alarm module is used for judging whether the gathering behavior is formed near the boundary of different areas.

[0040] The display module is used for displaying the gathering picture of the current gathering alarm.

[0041] Preferably, in the gathering behavior special alarm module, the specific judging mode is:

[0042] When the boundary of the key area in the high frequency period is reached, the total number v of the adjacent areas of the boundary is obtained, and the specific type area of the four type areas to which the v adjacent areas belong is determined, then the specific number of people near the boundary is collected every Tr time period, when the number of people collected for the first time exceeds L1 / v, it is marked as "1", when it is marked as "1" for three times in succession, the boundary of the v adjacent areas near the boundary of the key area in the high frequency period is determined, whether the number of people collected near the boundary of the corresponding adjacent area exceeds Lr / v is determined, when the number of people collected all exceeds Lr / v, it is determined that the gathering behavior is formed, the gathering alarm is sent out, and the shooting picture is sent to the subsequent display module, otherwise, no processing is performed.

[0043] Wherein, r=1, 2, 3, 4, which is selected according to the specific type area of the four type areas to which the adjacent areas belong: 1 represents the boundary of the key area in the high frequency period, 2 represents the boundary of the non-key area in the high frequency period, 3 represents the boundary of the key area in the low frequency period, and 4 represents the boundary of the non-key area in the low frequency period.

[0044] Similarly, whether the gathering behavior is formed in the boundaries of the other three areas is judged, and all the area boundaries are judged.

[0045] The present application provides a gathering alarm method based on personnel positioning scene.

[0046] (1) The method can collect the number of people in the target area in real time and accurately calculate the crowd density. By dividing a day into multiple periods, and further dividing the time in the period into hour periods, the method can more carefully analyze the number of people in each hour period, and determine the high frequency and low frequency periods based on the past data trend. This cumulative analysis of historical data helps to predict the future possible crowd gathering situation, and provides a scientific basis for event warning and resource allocation.

[0047] (2) The method subdivides the target area into multiple small areas through a region division model and dynamically determines key and non-key areas according to crowd density. This method can make the monitoring resources more concentrated in areas with higher crowd density, thereby improving monitoring efficiency and response speed. In addition, the algorithm can adjust the region division according to real-time data, which means that the system can adjust the monitoring focus in real time according to actual situation changes, rather than relying on preset division fixedly. Such flexibility is extremely advantageous for dealing with unpredictable gathering events.

[0048] (3) The method not only can judge the gathering behavior of each area, but also can issue an alarm when detecting the gathering behavior and send the related picture to the display module. This real-time alarm mechanism allows staff to understand the gathering situation in the initial stage and take timely measures to prevent potential safety risks. In emergency situations, quick and accurate response plays a crucial role in avoiding the expansion of the situation. At the same time, displaying the current gathering situation through visualization can help staff understand the on-site situation more intuitively, improving the efficiency and accuracy of decision-making. Moreover, the method does not need to rely on specific hardware devices, which greatly reduces the application cost and improves flexibility and popularity. BRIEF DESCRIPTION OF DRAWINGS

[0049] The present application will be further described below in conjunction with the accompanying drawings.

[0050] Figure 1 The framework diagram of the present application is as follows: DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0052] Embodiment one

[0053] Please refer to Figure 1 The present application provides a gathering alarm method based on personnel positioning scene, which includes:

[0054] The data acquisition module is used for collecting the number of people in the target area in real time, obtaining the specific area of the target area, and sending the number of people to the subsequent time period division module. The target area is the area that can be shot by multiple cameras.

[0055] The period division module is used for receiving the number of people information in real time, dividing a day into a plurality of time periods, determining the crowd density of each time period according to the number of people information, then selecting a time period as a target time period in turn, obtaining the crowd density of the target time period from n days before the current time, determining the target time period as a high-frequency time period or a low-frequency time period through a preset formula, so as to divide the plurality of time periods of a day into high-frequency time periods or low-frequency time periods.

[0056] The determination method of the high-frequency time period and the low-frequency time period comprises the following steps:

[0057] S1: dividing a day into y time periods, wherein y is a preset value, which is specifically set by a staff, and y=24 in the embodiment;

[0058] S2: selecting a time period as a target time period at random, dividing the target time period into x hour periods averagely, obtaining the specific number of people Rk in the target area at the end of each hour period according to the number of people information, then obtaining the specific area S of the target area, obtaining the crowd density Gk of the target time period at this time through calculation Gk=Rk / s, then obtaining the result U as the crowd density of the target time period through the formula wherein x is a preset value, which is specifically set by a staff, and 1≦k≦x;

[0059] S3: then obtaining the crowd density Ui of the target time period from n days before the current time, wherein n>1 and 1≦i≦n, and n=90 in the embodiment;

[0060] S4: then calculating the average value Up of the n groups of crowd density Ui, that is, S5: then calculating the deviation value H of the n groups of crowd density Ui, that is, S6: judging the size of the deviation value H:

[0061] S61: when H≤D1, taking the average value Up as the crowd density value Usp of the target time, and comparing the crowd density value Usp with a preset value D2, when Usp>D2, determining that the target time period is a high-frequency time period, and when Usp≦D2, determining that the target time is a low-frequency time period;

[0062] S62: when H>D1, it is determined that the deviation value H is too large, sorting Ui in the order of |Ui-Up| from large to small, deleting the first Ui each time and recalculating the deviation value H, until H≦D1;

[0063] then counting the number D3 of the remaining Ui, calculating the ratio of D3 to n, and comparing the ratio with a preset ratio D4:

[0064] When the ratio of D3 to n is greater than the preset ratio D4, the average value of the remaining Ui is calculated as the crowd density value Usp of the target period, and the crowd density value Usp is compared with the preset value D2. When Usp>D2, the target period is determined to be a high-frequency period, and when Usp≦D2, the target period is determined to be a low-frequency period;

[0065] When the ratio of D3 to n is less than or equal to the preset ratio D4, the average value of the deleted Ui is calculated as the crowd density value Usp at the target time, and the crowd density value Usp is compared with the preset value D2. When Usp>D2, the target time period is determined to be a high-frequency period. When Usp≦D2, the target time period is determined to be a low-frequency period.

[0066] Wherein, D1 and D2 are preset values, and the specific parameters of D1, D2 and D4 are set by the staff. In this embodiment, D4 is set to 0.55;

[0067] S7: Repeat steps S2-S6 to divide the y time periods into high-frequency time periods or low-frequency time periods;

[0068] It should be noted that a high-frequency period indicates that the number of people coming and going is relatively frequent during this period, and a low-frequency period indicates that the number of people coming and going is relatively sparse during this period.

[0069] The regional division module divides the target area into multiple small areas using a pre-set regional division model, determines an analysis period, sequentially uses the multiple small areas as analysis areas, and determines whether the analysis area is a key area or a non-key area, wherein the analysis period is determined in a high-frequency period or a low-frequency period;

[0070] The division of key areas into non-key areas includes the following steps:

[0071] P1: Divide the target area into m small areas on average, and each small area has the same area, denoted as and P2: Select a high-frequency period or, if there is no high-frequency period, select a period with the largest crowd density value Usp as the analysis period, and divide the analysis period into x hours of time.

[0072] P3: Randomly select a small area from the m small areas as the analysis area. At the end of each hour, obtain the specific number of people Rq in the analysis area, and then calculate the crowd density Gq=Rq / Sj in the analysis area at this moment. Then, use the formula Obtain the specific crowd density Zj of the analysis area in the analysis period, wherein, 1≦q≦x; P4: then repeat P3, obtain the specific crowd density Zj of all m small areas as analysis areas in the analysis period, then compare the crowd density Zj with the preset value D5, when the small area of Zj>D5 is determined as the key area, when the small area of Zj≦D5 is determined as the non-key area, wherein, D5 is the preset value, which is specifically set by the staff;

[0073] The area determination module is used for dividing the target area into four types of areas, i.e. the key area in the high-frequency period, the non-key area in the high-frequency period, the key area in the low-frequency period, and the non-key area in the low-frequency period according to the period division module and the area division module.

[0074] The gathering behavior alarm module is used for judging whether the gathering behavior appears in each type of area according to the four types of areas, and issuing the gathering alarm and sending the shooting picture to the subsequent display module when the gathering behavior is judged to appear.

[0075] Embodiment two

[0076] In the embodiment, on the basis of the embodiment one and different from the embodiment one, the embodiment mainly judges whether the gathering behavior appears in each type of area, including the following steps:

[0077] AS1: in the key area in the high-frequency period, the specific number of people is collected every T1 period of time in real time, when the number of people is collected more than L1 for the first time, it is marked as "1", when it is marked as "1" for three times continuously, it is judged as the gathering behavior, the gathering alarm is issued and the shooting picture is sent to the subsequent display module;

[0078] AS2: in the non-key area in the high-frequency period, the specific number of people is collected every T2 period of time in real time, when the number of people is collected more than L2 for the first time, it is marked as "1", when it is marked as "1" for three times continuously, it is judged as the gathering behavior, the gathering alarm is issued and the shooting picture is sent to the subsequent display module;

[0079] AS3: in the key area in the low-frequency period, the specific number of people is collected every T3 period of time in real time, when the number of people is collected more than L3 for the first time, it is marked as "1", when it is marked as "1" for three times continuously, it is judged as the gathering behavior, the gathering alarm is issued and the shooting picture is sent to the subsequent display module;

[0080] AS4: in the non-key area in the low-frequency period, the specific number of people is collected every T4 period of time in real time, when the number of people is collected more than L4 for the first time, it is marked as "1", when it is marked as "1" for three times continuously, it is judged as the gathering behavior, the gathering alarm is issued and the shooting picture is sent to the subsequent display module;

[0081] Wherein, T1, T2, T3 and T4 are preset values, and T4>T3>T2>T1, the specific parameters are set by the staff, L1, L2, L3, L4 are preset values, and L1>L2>L3>L4; the specific parameters are set by the staff.

[0082] Embodiment three

[0083] In the specific implementation process, on the basis of and different from embodiment one and embodiment two, this embodiment further comprises a display module.

[0084] Embodiment four

[0085] In the specific implementation process, on the basis of and different from embodiment one and embodiment two, this embodiment further comprises an aggregation behavior special alarm module and a display module;

[0086] The aggregation behavior special alarm module is used to determine whether the aggregation behavior is formed near the boundary of different regions, and the specific determination method is:

[0087] All region boundaries are determined, and the key region boundary in the high frequency period is taken as an example:

[0088] In the key region boundary in the high frequency period, the total number v of adjacent regions of the boundary is obtained, and the specific type region of the four types of regions to which the v adjacent regions belong is determined, then the number of people in the specific boundary is collected every Tr time period, when the number of people collected for the first time exceeds L1 / v, it is marked as "1", when it is marked as "1" for three times in a row, it is determined that the boundary near the key region boundary in the high frequency period in the v adjacent regions, the number of people collected in the corresponding adjacent region boundary is whether it exceeds Lr / v, when the corresponding number of people collected exceeds Lr / v, it is determined that the aggregation behavior is formed, the aggregation alarm is sent out and the shooting picture is sent to the subsequent display module, otherwise no processing is done, wherein r=1, 2, 3, 4, which is selected according to the specific type region of the four types of regions to which the adjacent regions belong: 1 represents the key region boundary in the high frequency period, 2 represents the non-key region boundary in the high frequency period, 3 represents the key region boundary in the low frequency period, and 4 represents the non-key region boundary in the low frequency period.

[0089] Similarly, whether the aggregation behavior is formed is determined for the region boundaries of the other three types of regions.

[0090] The display module is used to display the aggregation picture of the current aggregation alarm, so that the staff can understand the aggregation situation on site in the first time and determine whether a dangerous situation occurs.

[0091] Embodiment five

[0092] The embodiment contains all the implementation processes of the above three groups of embodiments in the specific implementation process.

[0093] Some data in the above formula are dimensionless for numerical calculation, and the contents not described in detail in the specification all belong to the prior art known to those skilled in the art.

[0094] The above embodiments are only used to illustrate the technical method of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present application.

Claims

1. A gathering alarm method based on personnel positioning scenario, characterized in that: include: The data acquisition module is used to collect the number of people in the target area in real time, obtain the specific area of ​​the target area, and send the number of people information to the subsequent time period division module, where the target area is the area that can be captured by multiple cameras; The time period division module is used to receive the number of people information in real time, divide a day into multiple time periods, determine the crowd density of each time period based on the number of people information, then select a time period as the target time period, obtain the crowd density of the target time period n days before the current time, and determine whether the target time period is a high-frequency time period or a low-frequency time period using a preset formula, thereby dividing the multiple time periods of a day into high-frequency time periods or low-frequency time periods, where n is a preset value; The regional division module divides the target area into multiple small areas using a pre-set regional division model, determines an analysis period, sequentially uses the multiple small areas as analysis areas, and determines whether the analysis area is a key area or a non-key area, wherein the analysis period is determined in a high-frequency period or a low-frequency period; The area determination module is used to divide the target area into four types of areas according to the time period division module and the area division module, namely, the key area of ​​the high-frequency period, the non-key area of ​​the high-frequency period, the key area of ​​the low-frequency period, and the non-key area of ​​the low-frequency period; The gathering behavior alarm module is used to determine whether gathering behavior occurs in each of the four types of areas. When it is determined that gathering behavior occurs, a gathering alarm is issued and the captured image is sent to the subsequent display module; The method for determining the high-frequency period and the low-frequency period in the period division module includes the following steps: S1: Divide a day into y periods, where y is a preset value; S2: Randomly select a time period as the target time period, divide the target time period into x hours, and at the end of each hour, obtain the specific number of people Rk in the target area based on the number of people information, and then obtain the specific area S of the target area. By calculating Gk=Rk / S, the crowd density Gk of the target time period at this moment is obtained, and then the formula is used to calculate the crowd density Gk. The result U is used as the crowd density of the target period, where x is a preset value and 1≦k≦x; S3: Then obtain the crowd density Ui of the target period n days before the current moment, where n>1, 1≦i≦n, and in this embodiment, n=90; S4: Then calculate the average value Up of n groups of population density Ui, that is S5: Then calculate the deviation value H of n groups of crowd density Ui, that is S6: Determine the size of the deviation value H; In step S6, it also includes: S61: When H≤D1, the average value Up is used as the crowd density value Usp at the target time, and the crowd density value Usp is compared with the preset value D2. When Usp>D2, the target time period is determined to be a high-frequency time period; when Usp≦D2, the target time period is determined to be a low-frequency time period; S62: When H>D1, it is determined that the deviation value H is too large, and Ui is sorted from large to small according to the calculated |Ui-Up|. Each time, the first-ranked Ui is deleted and the deviation value H is recalculated until H≦D1; Then count the number D3 of remaining Ui, calculate the ratio of D3 to n, and compare it with the preset ratio D4: When the ratio of D3 to n is greater than the preset ratio D4, the average value of the remaining Ui is calculated as the crowd density value Usp of the target period, and the crowd density value Usp is compared with the preset value D2. When Usp>D2, the target period is determined to be a high-frequency period, and when Usp≦D2, the target period is determined to be a low-frequency period; When the ratio of D3 to n is less than or equal to the preset ratio D4, the average value of the deleted Ui is calculated as the crowd density value Usp at the target time, and the crowd density value Usp is compared with the preset value D2. When Usp>D2, the target time period is determined to be a high-frequency period. When Usp≦D2, the target time period is determined to be a low-frequency period. Among them, D1 and D2 are preset values; After step S6, the method further includes: S7: Then repeat steps S2-S6 to divide the y time periods into high-frequency time periods or low-frequency time periods.

2. The method for gathering alarm based on personnel positioning scenario according to claim 1, characterized in that: In the area division module, the division of key areas and non-key areas includes the following steps: P1: Divide the target area into m small areas, each with the same area, denoted as Sj, 1≦j≦m, and P2: Select a high-frequency period or, if there is no high-frequency period, select a period with the largest crowd density value Usp as the analysis period, and divide the analysis period into x hours of time. P3: Randomly select a small area from the m small areas as the analysis area. At the end of each hour, obtain the specific number of people Rq in the analysis area, and then calculate the crowd density Gq = Rq / Sj in the analysis area at this moment. Then, use the formula The specific population density Zj of the analysis area during the analysis period is obtained, where 1≦q≦x.

3. The method for gathering alarm based on personnel positioning scenario according to claim 2, characterized in that: After step P3, the method further includes: P4: Then repeat P3 to obtain the specific crowd density Zj of all m small areas as the analysis area during the analysis period, and then compare the crowd density Zj with the preset value D5.

4. The method for gathering alarm based on personnel positioning scenario according to claim 2, characterized in that: In step P4, it also includes: When Zj>D5, the small area is determined as the key area, and when Zj≦D5, the small area is determined as the non-key area, where D5 is a preset value.

5. The method for gathering alarm based on personnel positioning scenario according to claim 2, characterized in that: In the gathering behavior alarm module, determining whether gathering behavior occurs in each type of area includes the following steps: AS1: In key areas during high-frequency periods, the number of people is collected in real time every T1 period. When the number of people collected for the first time exceeds L1, it is marked as "1". If it is marked as "1" three times in a row, it is determined to be a gathering behavior, a gathering alarm is issued, and the captured image is sent to the subsequent display module; AS2: In non-key areas during high-frequency periods, the number of people is collected in real time every T2 period. The first time the number of people collected exceeds L2, it is marked as "1". If it is marked as "1" three times in a row, it is determined to be a gathering behavior, a gathering alarm is issued, and the captured image is sent to the subsequent display module; AS3: In key areas during low-frequency periods, the number of people is collected in real time every T3 time period. When the number of people collected for the first time exceeds L3, it is marked as "1". If it is marked as "1" three times in a row, it is determined to be a gathering behavior, a gathering alarm is issued, and the captured image is sent to the subsequent display module; AS4: In non-key areas during low-frequency periods, the number of people is collected in real time every T4 time period. When the number of people collected for the first time exceeds L4, it is marked as "1". If it is marked as "1" three times in a row, it is determined to be a gathering behavior, a gathering alarm is issued, and the captured image is sent to the subsequent display module; Among them, T1, T2, T3 and T4 are preset values, and T4>T3>T2>T1, L1, L2, L3, L4 are preset values, and L1>L2>L3>L4.

6. The method for gathering alarm based on personnel positioning scenario according to claim 1, characterized in that: Also includes: A special alarm module for gathering behavior is used to determine whether a crowd is gathering near the boundaries of different areas; The display module is used to display the gathering picture of the current gathering alarm.

7. The method for gathering alarm based on personnel positioning scenario according to claim 6, characterized in that: In the special alarm module for gathering behavior, the specific judgment method is as follows: At the boundary of a key area during a high-frequency period, the total number v of adjacent areas at the boundary is obtained, and the specific types of areas of the four types of areas to which the v adjacent areas belong are determined. Then, the specific number of people at the boundary is collected in real time every Tr time period. When the number of people collected for the first time exceeds L1 / v, it is marked as "1". When it is marked as "1" three times in a row, the boundary of the v adjacent areas close to the key area during the high-frequency period is determined. The number of people at the corresponding adjacent area boundary exceeds Lr / v. If the number of people collected exceeds Lr / v, it is determined to be a gathering behavior, a gathering alarm is issued, and the captured image is sent to the subsequent display module. Otherwise, no processing is performed; Among them, r = 1, 2, 3, 4, is selected according to the specific type of area of ​​the four types of areas to which the adjacent area belongs: 1 represents the boundary of the key area during the high-frequency period, 2 represents the boundary of the non-key area during the high-frequency period, 3 represents the boundary of the key area during the low-frequency period, and 4 represents the boundary of the non-key area during the low-frequency period; Similarly, the boundaries of the other three areas are judged to see whether clustering behavior is formed, and all the area boundaries are judged.

Images