Scene monitoring method and device, electronic equipment and storage medium

Abstract

The present disclosure provides a scene monitoring method and device, electronic equipment and storage medium, the method comprising: acquiring monitoring video collected by a monitoring device arranged at at least one monitoring point; determining whether a monitoring event occurs in a monitoring area corresponding to the at least one monitoring point based on the monitoring video; in the case where a monitoring event occurs in the monitoring area corresponding to the at least one monitoring point, acquiring people monitoring data matching the monitoring event within a preset time period; and determining people flow state data of the at least one monitoring device based on the people monitoring data matching the monitoring event within the preset time period.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of computer vision, and in particular, to a scene monitoring method and device, an electronic device, and a storage medium. BACKGROUND

[0002] With the improvement of people's living standards, more and more large-scale activities are held in various places. Because of the large number of people during the large-scale activities, accidents such as stampede and congestion are prone to occur in the places where the large-scale activities are held. Therefore, in order to ensure the safety of various places, effective monitoring of the flow of people is becoming more and more important. SUMMARY

[0003] Therefore, the present disclosure at least provides a scene monitoring method, device, electronic device, and storage medium.

[0004] In a first aspect, the present disclosure provides a scene monitoring method, comprising:

[0005] obtaining a monitoring video collected by a monitoring device arranged at at least one monitoring point;

[0006] determining whether a monitoring event occurs in a monitoring area corresponding to the at least one monitoring point based on the monitoring video;

[0007] in a case where the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point, obtaining people monitoring data matching the monitoring event in a preset time period;

[0008] determining people flow state data of the at least one monitoring device based on the people monitoring data matching the monitoring event in the preset time period.

[0009] In the above method, by obtaining the monitoring video collected by the monitoring device, in a case where it is detected that the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the collected monitoring video, the people monitoring data matching the monitoring event in the preset time period is obtained, and the people flow state data of the at least one monitoring device is determined based on the people monitoring data matching the monitoring event in the preset time period. The people flow state data represents the state of the monitoring event, and the monitoring of the monitoring video is realized. For example, the people flow state data can be the total number of people, and in a case where the total number of people is large, it is indicated that the monitoring event occurs frequently.

[0010] In a possible implementation, after the people flow state data of the at least one monitoring device is determined, the method further comprises:

[0011] in a case where the people flow state data of the at least one monitoring device meets an alarm condition, generating people flow state alarm information.

[0012] Herein, when the determined people flow state data satisfies the alarm condition, people flow state alarm information is generated, and based on the generated people flow state alarm information, the target monitoring area can be regulated to avoid the occurrence of safety accidents and ensure the safety of people flow in the target monitoring area.

[0013] In a possible implementation, before determining whether a monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the monitoring video, the method further includes:

[0014] generating a monitoring mark corresponding to the monitoring video and matching the monitoring event, wherein the monitoring mark includes at least one of the following: an in-out boundary line, an in direction, an out direction, and a polygon structure; and / or

[0015] For any video frame in the monitoring video, in response to a triggered human body labeling operation, human body bounding box information of a plurality of pedestrians located at different depth positions in the video frame is determined, wherein the human body bounding box information includes area information of the human body bounding box and depth information thereof.

[0016] In a possible implementation, the generating of the monitoring mark corresponding to the monitoring video and matching the monitoring event includes:

[0017] obtaining a video picture screenshot, wherein the video picture screenshot includes a pre-drawn monitoring mark matching the monitoring event;

[0018] determining position information of the monitoring mark in the video picture screenshot;

[0019] based on the position information corresponding to the monitoring mark, generating a monitoring mark matching the monitoring event in a video picture of the monitoring video.

[0020] In a possible implementation, the monitoring event includes an over-density event, and the determining of whether a monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the monitoring video includes:

[0021] For the monitoring area corresponding to each monitoring point, based on human body bounding box information marked in the human body labeling operation, a predicted area of the monitoring area corresponding to the monitoring mark is determined;

[0022] based on the detected number of people in the monitoring area and the predicted area, a people density corresponding to the monitoring area is determined;

[0023] in a case where the people density is greater than a set value, it is determined that the over-density event occurs in the monitoring area corresponding to the at least one monitoring point.

[0024] In a possible implementation, in a case where the monitoring event is a cross-line event, the method further includes:

[0025] determining, based on the monitoring video, whether a target object that crosses a target position matched with the pre-drawn in-out boundary exists in the monitoring area corresponding to the at least one monitoring point;

[0026] if the target object exists, determining that the cross-line event occurs in the monitoring area corresponding to the at least one monitoring point.

[0027] In the above implementation, in a case where the target object that crosses the target position matched with the in-out boundary exists in the monitoring area corresponding to the at least one monitoring point based on the monitoring video, it is determined that the cross-line event occurs in the monitoring area corresponding to the at least one monitoring point, real-time monitoring of the cross-line event is implemented, and the accuracy of cross-line event monitoring is improved.

[0028] In a possible implementation, in a case where the monitoring event is a cross-line event, the method further includes:

[0029] obtaining the number of people monitoring data matched with the monitoring event in a preset time period, including:

[0030] In the above method, in a case where the monitoring event is a cross-line event, the number of people flowing in and the number of people flowing out at different collection time points in a preset time period can be obtained, which provides data support for subsequent determination of people flow state data corresponding to the cross-line event.

[0031] In a possible implementation, in a case where the monitoring point is one, the method further includes:

[0032] determining, based on the number of people flowing in and the number of people flowing out at different collection time points in a preset time period, the total number of people flowing in and the total number of people flowing out in the monitoring area corresponding to the monitoring point in the preset time period;

[0033] In a case where the people flow state data of the at least one monitoring device meets an alarm condition, generating people flow state alarm information, including:

[0034] generate the crowd state alarm information in a case where the total number of people entering the monitoring area in the preset time period is greater than a first threshold of the number of people entering the monitoring area, and / or in a case where the total number of people leaving the monitoring area in the preset time period is greater than a second threshold of the number of people leaving the monitoring area.

[0035] Here, in a case where the monitoring point is one, the total number of people entering the monitoring area in the preset time period and the total number of people leaving the monitoring area in the preset time period in the monitoring area corresponding to the monitoring point are determined based on the number of people entering and the number of people leaving at different collection time points in the preset time period. In a case where the total number of people entering the monitoring area in the preset time period is greater than a first threshold of the number of people entering the monitoring area, and / or in a case where the total number of people leaving the monitoring area in the preset time period is greater than a second threshold of the number of people leaving the monitoring area, the crowd state alarm information is generated, realizing early warning of the number of people entering and the number of people leaving the monitoring video, so as to guide the crowd based on the generated crowd state alarm information, avoiding the occurrence of safety accidents caused by a large number of people entering or a large number of people leaving in a short time.

[0036] In a possible implementation, in a case where the monitoring point is one, the crowd state data of the at least one monitoring device is determined based on the number of people monitoring data matching the monitoring event in the preset time period, including:

[0037] The speed of people entering and the speed of people leaving in the monitoring area corresponding to the monitoring point are determined based on the number of people entering and the number of people leaving at different collection time points in the preset time period.

[0038] In the above method, the speed of people entering and the speed of people leaving in the monitoring area corresponding to the monitoring point can be determined based on the number of people entering and the number of people leaving at different collection time points in the preset time period, realizing monitoring of the speed of people entering and the speed of people leaving, and avoiding the occurrence of safety accidents caused by a large speed of people entering or a large speed of people leaving.

[0039] In a possible implementation, in a case where the monitoring point is one, the crowd state data of the at least one monitoring device is determined based on the number of people monitoring data matching the monitoring event in the preset time period, including:

[0040] For each monitoring point, the total number of people entering the monitoring area in the preset time period and the total number of people leaving the monitoring area in the preset time period in the monitoring area corresponding to the monitoring point are determined based on the number of people entering and the number of people leaving at different collection time points in the preset time period.

[0041] The net amount of people in the target monitoring area is determined based on the historical number of people of the target monitoring area in the preset time period and the total number of people entering and the total number of people leaving in the preset time period corresponding to the plurality of monitoring points.

[0042] generate the people flow state alarm information in a case where it is determined that the people flow state data of the at least one monitoring device meets an alarm condition, including:

[0043] generate the people flow state alarm information in a case where it is determined that the net population in the target monitoring area is greater than a set net population threshold.

[0044] Here, after determining the total in-flow quantity and the total out-flow quantity in the preset time period in the monitoring area corresponding to each monitoring point, the net population in the target monitoring area can be determined based on the historical population of the target monitoring area in the preset time period and the total in-flow quantity and the total out-flow quantity of the preset time period corresponding to the plurality of monitoring points respectively, and the people flow state alarm information is generated in a case where the net population in the target monitoring area is greater than a set net population threshold, realizing early warning of the net population in the target monitoring area, so as to guide the personnel based on the generated people flow state alarm information when the net population is large, avoiding the occurrence of safety accidents when the target monitoring area has a large number of people.

[0045] In a possible implementation, in a case where the monitoring event is an overpopulation event, determining whether the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the monitoring video includes:

[0046] determining whether the number of target objects in the monitoring area corresponding to the at least one monitoring point exceeds an overpopulation threshold based on the monitoring video;

[0047] if yes, determining that the overpopulation event occurs in the monitoring area corresponding to the at least one monitoring point.

[0048] In the above method, when it is determined that the number of target objects in the monitoring area corresponding to the at least one monitoring point exceeds the overpopulation threshold based on the monitoring video, it is determined that the overpopulation event occurs in the monitoring area corresponding to the at least one monitoring point, realizing real-time monitoring of the overpopulation event and improving the accuracy of overpopulation event monitoring.

[0049] In a possible implementation, in a case where the monitoring event is an overpopulation event, obtaining population monitoring data matching the monitoring event in a preset time period includes:

[0050] counting the number of target objects at different collection time points in the preset time period.

[0051] In the above method, when the monitoring event is an overpopulation event, the number of target objects at different collection time points in the preset time period can be counted, providing data support for subsequent determination of people flow state data corresponding to the overpopulation event.

[0052] In a possible implementation, when the monitoring point is one, the people flow state data of the at least one monitoring device is determined based on the number of people matching the monitoring event in a preset time period, and the people flow state data includes:

[0053] The average number of people in the monitoring area corresponding to the monitoring point in the preset time period is determined based on the number of target objects at different collection time points in the preset time period.

[0054] When the people flow state data of the at least one monitoring device meets an alarm condition, people flow state alarm information is generated, and the people flow state alarm information includes:

[0055] When the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is greater than a first number threshold, the people flow state alarm information is generated.

[0056] In the method, when the monitoring point is one, the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is determined based on the number of target objects at different collection time points in the preset time period, and when the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is greater than a first number threshold, the people flow state alarm information is generated, so that the average number of people in the detection area of the monitoring video is monitored, and the detection area is guided based on the generated people flow state alarm information, to avoid safety accidents caused by a large number of people in the detection area.

[0057] In a possible implementation, when the monitoring point is one, the people flow state data of the at least one monitoring device is determined based on the number of people matching the monitoring event in a preset time period, and the people flow state data includes:

[0058] For each monitoring point, the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is determined based on the number of target objects at different collection time points in the preset time period.

[0059] The total real-time number of people in the target monitoring area is determined based on the average number of people corresponding to each of the plurality of monitoring points.

[0060] When the people flow state data of the at least one monitoring device meets an alarm condition, people flow state alarm information is generated, and the people flow state alarm information includes:

[0061] When the total real-time number of people in the target monitoring area is greater than a second number threshold, the people flow state alarm information is generated.

[0062] In the method, after determining the average number of people in the monitoring area corresponding to each monitoring point in the preset time period, the total real-time number of people in the target monitoring area can be determined based on the average number of people corresponding to each of the plurality of monitoring points. If the total real-time number of people in the target monitoring area is greater than the second number threshold, the flow state alarm information is generated, thereby realizing early warning of the total real-time number of people in the target monitoring area. When the total real-time number of people is large, the generated flow state alarm information can be used to guide the personnel in the target monitoring area, thereby avoiding the occurrence of safety accidents when the total real-time number of people in the target monitoring area is large.

[0063] In a possible implementation, the method further includes:

[0064] averaging the flow state data at the same collection time point in the plurality of historical dates to obtain predicted flow state data corresponding to each collection time point;

[0065] The predicted flow state data corresponding to each collection time point constitutes predicted data of the flow state data in the future date; and the predicted data is used to generate a flow guidance plan.

[0066] The effects of the following devices, electronic devices, and the like are described in the above method, which will not be repeated here.

[0067] In a second aspect, the present disclosure provides a scene monitoring device, comprising:

[0068] A first acquisition module is configured to acquire monitoring videos collected by monitoring devices arranged at at least one monitoring point.

[0069] A detection module is configured to determine whether a monitoring event occurs in a monitoring area corresponding to the at least one monitoring point based on the monitoring videos.

[0070] A second acquisition module is configured to acquire people monitoring data matching the monitoring event in a preset time period if the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point.

[0071] A determination module is configured to determine flow state data of the at least one monitoring device based on the people monitoring data matching the monitoring event in the preset time period.

[0072] In a third aspect, the present disclosure provides an electronic device, comprising a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate through the bus. The machine-readable instructions are executed by the processor to perform the steps of the scene monitoring method according to the first aspect or any of the embodiments.

[0073] In a fourth aspect, the present disclosure provides a computer readable storage medium, having stored thereon a computer program, which, when executed by a processor, performs the steps of the scene monitoring method according to the first aspect or any one of the embodiments.

[0074] In order to make the above objectives, characteristics and advantages of the present disclosure more obvious and easy to understand, the following preferred embodiments are specifically described below, and the accompanying drawings are described in detail as follows. BRIEF DESCRIPTION OF DRAWINGS

[0075] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following will briefly introduce the drawings needed to be used in the embodiments. The drawings herein are incorporated into the specification and form a part of the specification, which show the embodiments consistent with the present disclosure, and are used to explain the technical solutions of the present disclosure together with the specification. It should be understood that the following drawings only show some embodiments of the present disclosure, and therefore should not be considered as a limitation to the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor.

[0076] Figure 1 A flowchart of a scene monitoring method provided by an embodiment of the present disclosure is shown;

[0077] Figure 2a An interface diagram showing a screenshot of a video picture provided by an embodiment of the present disclosure is shown;

[0078] Figure 2b An interface diagram showing a screenshot of a video picture provided by an embodiment of the present disclosure is shown;

[0079] Figure 3 An interface diagram showing detailed information of a crowd state alarm provided by an embodiment of the present disclosure is shown;

[0080] Figure 4a An interface diagram showing detailed information of a crowd state alarm provided by an embodiment of the present disclosure is shown;

[0081] Figure 4b An interface diagram showing detailed information of a crowd state alarm provided by another embodiment of the present disclosure is shown;

[0082] Figure 4c An interface diagram showing alarm details provided by an embodiment of the present disclosure is shown;

[0083] Figure 5 An interface diagram showing detailed information of a crowd state alarm provided by an embodiment of the present disclosure is shown;

[0084] Figure 6a Fig. 1 shows a detailed interface diagram for displaying a flow state alarm according to an embodiment of the present disclosure;

[0085] Figure 6b Fig. 2 shows a detailed interface diagram for displaying alarm details according to an embodiment of the present disclosure;

[0086] Figure 7 Fig. 3 shows an architecture diagram of a scene monitoring device according to an embodiment of the present disclosure;

[0087] Figure 8 Fig. 4 shows a structural diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0088] In order to make the objectives, technical solutions and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. The components of the embodiments of the present disclosure described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present disclosure provided in the drawings is not intended to limit the scope of the claimed present disclosure, but only represents selected embodiments of the present disclosure. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present disclosure.

[0089] With the improvement of people's living standards, more and more large-scale activities are held in various places. Because of the large number of people during the large-scale activities, accidents such as stampede and congestion are prone to occur in the places where the large-scale activities are held. Therefore, in order to ensure the safety of various places, effective monitoring of people flow is becoming more and more important. In order to solve the above problems and improve the safety of places, the embodiments of the present disclosure provide a scene monitoring method and device, an electronic device and a storage medium.

[0090] To make the present disclosure embodiments easy to understand, first, a scene detection method disclosed by the present disclosure embodiments is introduced in detail. The execution subject of the scene monitoring method provided by the present disclosure embodiments is generally a computer device with certain computing capability, which for example includes a terminal device or a server or other processing device, and the terminal device can be a user equipment (User Equipment, UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital processing (Personal Digital Assistant, PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. In some possible implementation manners, the scene monitoring method can be realized by a processor calling computer readable instructions stored in a memory.

[0091] Referring to Figure 1 Fig. 1 is a flowchart of a scene monitoring method provided by the present disclosure embodiments, and the method includes S101-S104, and specifically:

[0092] S101, acquiring monitoring videos collected by monitoring devices arranged at at least one monitoring point.

[0093] S102, determining whether a monitoring event occurs in a monitoring area corresponding to the at least one monitoring point based on the monitoring videos.

[0094] S103, in a case where the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point, acquiring people number monitoring data matching the monitoring event in a preset time period.

[0095] S104, determining people flow state data of the at least one monitoring device based on the people number monitoring data matching the monitoring event in the preset time period.

[0096] In the above method, by acquiring the monitoring videos collected by the monitoring devices, in a case where it is detected based on the collected monitoring videos that the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point, the people number monitoring data matching the monitoring event in the preset time period is acquired, and the people flow state data of the at least one monitoring device is determined based on the people number monitoring data matching the monitoring event in the preset time period, the state of the monitoring event is represented by the determined people flow state data, and the monitoring of the monitoring videos is realized, for example, the people flow state data can be the total number of people flow, and in a case where the total number of people flow is large, it is represented that the monitoring event occurs frequently.

[0097] The following specifically describes S101-S104.

[0098] For S101:

[0099] In implementation, the method can be used to detect a target monitoring area, which can be any area in a real scene, such as a shopping mall, a beach, a park, a subway station, etc.

[0100] For example, a plurality of monitoring points can be set at the target monitoring area, and a monitoring device is installed at each monitoring point, so that the monitoring device can monitor the corresponding monitoring area, and the monitoring of the target monitoring area is realized. The setting of the monitoring points can be determined according to actual needs, for example, when the target monitoring area is a shopping mall, a monitoring point can be set at each door of the shopping mall and / or a monitoring point can be set at each elevator entrance, etc.

[0101] In implementation, the monitoring device can be a monitoring camera or the like. By setting a monitoring device at each monitoring point, the monitoring video in the corresponding monitoring area is collected by the monitoring device, so that the monitoring video collected by each monitoring device can be obtained, i.e., the monitoring video collected by the monitoring device set at at least one monitoring point can be obtained.

[0102] For S102:

[0103] Here, for the monitoring video collected by each monitoring device, it is determined whether a monitoring event occurs in the monitoring area corresponding to the monitoring point based on the monitoring video, and then it is determined whether a monitoring event occurs in the monitoring area corresponding to each monitoring point in the at least one monitoring point. The monitoring event can include an over-crowding event and / or a cross-line event. The over-crowding event refers to that the number density in the area is greater than the set value, i.e., the personnel density in the area is large. The cross-line refers to that a pedestrian in the area crosses a set reference line.

[0104] The cross-line event and the over-crowding event are described in combination with specific scenarios, for example, for the cross-line event, in a subway station, a reference line can be set at a position on the platform at a preset distance (e.g., 1 meter) from the subway, and it is monitored whether anyone crosses the reference line (i.e., whether anyone crosses the reference line to enter or exit the subway), and if so, a cross-line event occurs. For example, for the over-crowding event, on a beach, a target monitoring area can be set, and when the number of people in the target monitoring area is greater than the set number of people, it is determined that an over-crowding event occurs.

[0105] In another optional implementation, a function button can also be set for each monitoring video, and the monitoring of the monitoring event in the monitoring area of the monitoring point is determined by triggering the function button. For example, a first function button corresponding to the cross-line event (monitoring of the cross-line event for a single monitoring video) can be set, and after the first function button corresponding to the monitoring video A is triggered, the monitoring of the cross-line event for the monitoring video A is determined. Alternatively, a second function button corresponding to the cross-line event (monitoring of the cross-line event for a monitoring video group) can also be set, and after the second function button corresponding to the monitoring video group A composed of the monitoring video A, the monitoring video B, etc. is triggered, the monitoring of the cross-line event for the monitoring video group A is determined.

[0106] For another example, a third function button corresponding to the over-density event (monitoring of the over-density event for a single monitoring video) can also be set, and after the third function button corresponding to the monitoring video A is triggered, the monitoring of the over-density event for the monitoring video A is determined. Alternatively, a fourth function button corresponding to the over-density event (monitoring of the over-density event for a monitoring video group) can also be set, and after the fourth function button corresponding to the monitoring video group A composed of the monitoring video A, the monitoring video B, etc. is triggered, the monitoring of the over-density event for the monitoring video group A is determined.

[0107] In specific implementation, before determining whether the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the monitoring video, the monitoring mark corresponding to the monitoring video can be drawn. For the cross-line event, the monitoring mark can be the pre-drawn in-out boundary, in direction and out direction; for the over-density event, the monitoring mark can be any pre-drawn polygon, or for the over-density event, the corresponding monitoring mark can not be set. The monitoring marks corresponding to different monitoring videos are different, that is, for each monitoring video, the corresponding monitoring mark (the monitoring mark corresponding to the cross-line event and / or the monitoring mark corresponding to the over-density event) can be drawn for the monitoring video.

[0108] In implementation, for each monitoring video, a video screenshot can be collected from the monitoring video, and the video screenshot is displayed, so that the user can draw a monitoring mark on the video screenshot according to actual needs. Then, the video screenshot with the pre-drawn monitoring mark is acquired, and the position information of the monitoring mark in the video screenshot is determined, where the position information can be a coordinate set of the monitoring mark in a pixel coordinate system corresponding to the video screenshot, such as the position information of the in-out boundary. Further, the target position information matched with the monitoring mark can be determined in the video screenshot of the monitoring video. When the installation information such as the position and orientation of the monitoring device does not change, the position information of the monitoring mark in the video screenshot can be the target position information of the monitoring mark in the video screenshot of the monitoring video. Further, whether a monitoring event occurs in the monitoring area corresponding to the monitoring point can be determined based on the monitoring video collected for the monitoring device and the determined target position information.

[0109] For example, when the monitoring mark includes the monitoring mark corresponding to the cross-line event, referring to a schematic diagram of an interface for displaying the video screenshot with the drawn monitoring mark shown in FIG. 21, the interface includes the pre-drawn monitoring mark 21, which includes the drawn in-out boundary and the arrow mark indicating the in-out direction. Figure 2a The interface includes the pre-drawn monitoring mark 21, which includes the drawn in-out boundary and the arrow mark indicating the in-out direction. Figure 2a In the drawing, the in-out boundary and the arrow mark indicating the in-out direction are drawn on the video screenshot. In addition, the in-person flow threshold (i.e., the first in-person flow threshold) and / or the out-person flow threshold (i.e., the second in-person flow threshold) can be set on the displayed interface, so as to monitor the monitoring video based on the set in-person flow threshold and / or the out-person flow threshold. The interface also includes the prompt information of the cross-line event setting located above the video screenshot, so that the user can draw the monitoring mark according to the displayed prompt information of the cross-line event setting. When the monitoring mark is drawn, the “redraw” button can be triggered to delete the drawn monitoring mark and redraw a new monitoring mark.

[0110] For example, when the monitoring mark includes the monitoring mark corresponding to the cross-line event, referring to a schematic diagram of an interface for displaying the video screenshot with the drawn monitoring mark shown in FIG. 21, the interface includes the pre-drawn monitoring mark 21, which includes the drawn in-out boundary and the arrow mark indicating the in-out direction. Figure 2bAn interface diagram for drawing a video screenshot with a monitoring mark is shown, which includes a pre-drawn monitoring mark 21, and the monitoring mark includes a polygon indicating a detection area, wherein the number of detection areas can be multiple. When drawing the monitoring mark, the number of people in the early warning can also be set on the displayed interface, that is, the number of people in the early warning corresponding to the general risk, the number of people in the early warning corresponding to the greater risk, and the number of people in the early warning corresponding to the major risk, so as to monitor the monitoring video based on the set number of people in the early warning. The figure also includes prompt information of the over-concentration event setting above the video screenshot, so that the user can draw the monitoring mark indicating the detection area according to the displayed over-concentration event setting prompt information. When drawing the monitoring mark, the "redraw" button can also be triggered to delete the drawn monitoring mark and redraw a new monitoring mark. After drawing the monitoring mark corresponding to the monitoring video, the drawn monitoring mark can be stored in the reuse area, so that the next time the monitoring mark is determined, the function button of the reuse area can be directly triggered to realize the reuse of the monitoring mark.

[0111] Figure 2b The function button of the human body annotation in the interface diagram is used to display the setting information of the human body annotation. It is considered that the area size of the human body in the monitoring video screenshot is related to the height and angle of the monitoring device, and the distance between the same human body and the monitoring device is different, and the area size in the monitoring video screenshot is different, that is, when the distance to the monitoring device is closer, the area of the human body is larger, so the human body annotation is the basis setting of the over-concentration event and the over-concentration event.

[0112] Specifically, in the video screenshot, a plurality of human body frames of pedestrians located at different depth positions are marked, the area of each human body frame of the pedestrian and the depth information thereof are estimated; the human body annotation result is used to facilitate the algorithm (such as an image recognition algorithm for recognizing human bodies) to recognize human bodies in different monitoring devices under different conditions, and improve the recognition accuracy, wherein the more human body frames, the higher the accuracy, and in specific implementation, the number of marked human body frames can be set as needed, such as setting the number of marked human body frames to be in the range of 3-10. Further, the area of the plurality of human body frames of the pedestrians and the depth information of each pedestrian can be used to detect the real-time number of people included in the detection area of each second video screenshot in the monitoring video.

[0113] In the over-concentration event, when the detection area (monitoring mark) is drawn, the predicted area of the drawn area in the real scene can be calculated based on the human body samples marked in the human body annotation, and the predicted area is displayed at the "area area estimation" below Figure 2b , and the personnel density in the detection area can be calculated in subsequent point over-concentration alarm, video group over-concentration alarm, etc. In addition, the Figure 2bThe middle also includes a "correct area" function button, which can correct the predicted area displayed on the area estimation after triggering the "correct area" function button.

[0114] In an optional embodiment, when the monitoring event is a cross-line event, the monitoring video is used to determine whether the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point, including:

[0115] The monitoring video is used to determine whether there is a target object crossing the target position matched with the pre-drawn in-out boundary in the monitoring area corresponding to the at least one monitoring point.

[0116] If yes, it is determined that the cross-line event occurs in the monitoring area corresponding to the at least one monitoring point.

[0117] When the monitoring event is a cross-line event, the monitoring video collected for each monitoring point is used to determine whether there is a target object crossing the target position matched with the in-out boundary in the monitoring area corresponding to the monitoring point, such as detecting whether a pedestrian crosses the drawn in-out boundary in the monitoring video. If yes, it is determined that the cross-line event occurs in the monitoring area corresponding to the monitoring point. If no, it is determined that the cross-line event does not occur in the monitoring area corresponding to the monitoring point.

[0118] The monitoring area corresponding to the monitoring point can be a detection area that can be monitored by the monitoring device arranged at the monitoring point. The monitoring area corresponding to the monitoring point is related to the installation position and installation angle of the monitoring device, and different installation positions and installation angles correspond to different monitoring areas.

[0119] In the above embodiment, when the monitoring video is used to determine that there is a target object crossing the target position matched with the in-out boundary in the monitoring area corresponding to the at least one monitoring point, it is determined that the cross-line event occurs in the monitoring area corresponding to the at least one monitoring point, which realizes real-time monitoring of the cross-line event and improves the accuracy of the cross-line event monitoring.

[0120] In an optional embodiment, when the monitoring event is an over-density event, the monitoring video is used to determine whether the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point, including:

[0121] The monitoring video is used to determine whether the number of target objects in the monitoring area corresponding to the at least one monitoring point exceeds an over-density threshold.

[0122] If yes, it is determined that the over-density event occurs in the monitoring area corresponding to the at least one monitoring point.

[0123] When the monitoring event is the over-density event, for the monitoring video collected at each monitoring point, the number of target objects in the monitoring area corresponding to the monitoring point can be determined based on the monitoring video, whether the number of people in the monitoring area is greater than the over-density threshold of the social group, if yes, it is determined that the over-density event occurs in the monitoring area corresponding to the monitoring point; if no, it is determined that the over-density event does not occur in the monitoring area corresponding to the monitoring point. Here, when the monitoring event is the over-density event, the monitoring area corresponding to the monitoring point can be the detection area matched with the drawn polygon; when the monitoring mark is not drawn, the monitoring area corresponding to the monitoring point is the detection area that the monitoring device set at the monitoring point can monitor (i.e. the area corresponding to the monitoring interface of the monitoring video is the monitoring area).

[0124] In the above method, when it is determined that the number of target objects in the monitoring area corresponding to at least one monitoring point exceeds the over-density threshold based on the monitoring video, it is determined that the over-density event occurs in the monitoring area corresponding to at least one monitoring point, which realizes real-time monitoring of the over-density event and improves the accuracy of over-density event monitoring.

[0125] For S103 and S104:

[0126] Here, when it is determined that the monitoring event occurs in the monitoring area corresponding to at least one monitoring point, the number of people monitoring data matched with the monitoring event in a preset time period can be obtained; the number of people monitoring data includes the number of people monitoring data corresponding to the cross-line event and / or the number of people monitoring data corresponding to the over-density event. Further, based on the number of people monitoring data matched with the monitoring event in the preset time period, the people flow state data of at least one monitoring device is determined; the people flow state data includes the people flow state data corresponding to the cross-line event and / or the people flow state data corresponding to the over-density event. The preset time period can be set as needed, for example, the preset time period can be the time period from the time when the monitoring event is determined to occur to one hour later, if the time when the monitoring event is determined to occur is 13:10:00, the preset time period is the time period from 13:10:00 to 14:10:00. For example, the preset time period can be the time period from the time when the monitoring event is determined to occur to one minute later, if the time when the monitoring event is determined to occur is 13:10:00, the preset time period is the time period from 13:10:00 to 13:11:00.

[0127] For the cross-line event, the number of people monitoring data matched with the cross-line event in a preset time period can be obtained; and based on the number of people monitoring data matched with the cross-line event in the preset time period, the people flow state data matched with the cross-line event of at least one monitoring device is determined.

[0128] For the over-dense event, the number of people monitoring data in the preset time period matched with the over-dense event can be acquired; and based on the number of people monitoring data in the preset time period matched with the over-dense event, the people flow state data of the at least one monitoring device matched with the over-dense event is determined.

[0129] In an optional implementation, after the people flow state data of the at least one monitoring device is determined, the method further includes: in a case where the people flow state data of the at least one monitoring device meets an alarm condition, generating people flow state alarm information.

[0130] Here, it is determined whether the people flow state data of the at least one monitoring device meets the alarm condition, and if so, the people flow state alarm information is generated, so that the user can generate a relief plan based on the people flow state alarm information to avoid events such as stampede and congestion in the target monitoring area.

[0131] Here, when the determined people flow state data meets the alarm condition, the people flow state alarm information is generated, and based on the generated people flow state alarm information, the target monitoring area can be regulated to avoid the occurrence of safety accidents and ensure the safety of people flow in the target monitoring area.

[0132] The following describes the alarm process of the cross-line event and the alarm process of the over-dense event respectively.

[0133] First, the alarm process of the cross-line event is described.

[0134] In a case where the monitoring event is a cross-line event, the number of people monitoring data in the preset time period matched with the monitoring event is acquired, including:

[0135] The number of in-flowing people and the number of out-flowing people at different collection time points in the preset time period are acquired, wherein the number of in-flowing people at different collection time points refers to the number of people crossing the pre-drawn in-out boundary in the pre-drawn in direction at different collection time points; and the number of out-flowing people at different collection time points refers to the number of people crossing the pre-drawn in-out boundary in the pre-drawn out direction at different collection time points.

[0136] Here, the monitoring identifier corresponding to the cross-line event can include the pre-set in-out boundary and in-out direction (in direction and / or out direction, the out direction being the opposite direction of the in direction), the in-out boundary can divide the monitoring area corresponding to the monitoring video into an in area and an out area, the in direction in the in-out direction can be the direction from the out area to the in area, and the out direction in the in-out direction can be the direction from the in area to the out area.

[0137] Further, the number of people entering and the number of people leaving at each collection time point in the preset time period in the monitoring video can be determined based on the set entry-exit boundary, the entry-exit direction and the monitoring video. The number of people entering at different collection time points refers to the number of people crossing the entry-exit boundary in the entry direction at different collection time points. The number of people leaving at different collection time points refers to the number of people crossing the entry-exit boundary in the exit direction at different collection time points.

[0138] For example, the trained target tracking algorithm can be used to detect the monitoring video based on the set monitoring identifier. The detection result is output once every preset time in the preset time period. The multiple detection results in the preset time period can be the number of people entering and the number of people leaving at different collection time points in the preset time period. Each detection result is associated with an output time (which is the collection time point). Thus, the number of people entering and the number of people leaving at different collection time points in the preset time period can be obtained.

[0139] In the above method, when the monitoring event is a crossing event, the number of people entering and the number of people leaving at different collection time points in the preset time period can be obtained, which provides data support for determining the people flow state data corresponding to the crossing event.

[0140] In an optional implementation, when the monitoring point is one, the people flow state data of the at least one monitoring device is determined based on the number of people monitoring data matching the monitoring event in the preset time period, including: determining the total number of people entering and the total number of people leaving in the monitoring area corresponding to the monitoring point in the preset time period based on the number of people entering and the number of people leaving at different collection time points in the preset time period.

[0141] When the people flow state data of the at least one monitoring device meets the alarm condition, the people flow state alarm information is generated, including: when it is determined that the total number of people entering in the preset time period is greater than the set first people flow threshold and / or the total number of people leaving in the preset time period is greater than the set second people flow threshold, the people flow state alarm information is generated.

[0142] After obtaining the number of people monitoring data matching the monitoring event (crossing event) in the preset time period, i.e., for the crossing event, after obtaining the number of people entering and the number of people leaving at different collection time points in the preset time period, the total number of people entering and the total number of people leaving in the monitoring area corresponding to the monitoring point in the preset time period can be determined based on the number of people entering and the number of people leaving at different collection time points in the preset time period.

[0143] With the above embodiment continuing to illustrate, the trained target tracking algorithm can output a detection result every 3 seconds (determine a collection time point every 3 seconds), and the detection result can be the number of people in and the number of people out in the 3 seconds, for example, the detection result can be: the number of people in between 08:10:01-08:10:03 (including 10:01 and 10:03) is 20, the number of people out is 50, and the associated output time (collection time point) is 08:10:03; and then multiple detection results in the preset time period can be obtained, that is, the number of people in and the number of people out at different collection time points in the preset time period.

[0144] After obtaining the number of people in and the number of people out at different collection time points in the preset time period, the number of people in at different collection time points can be added to obtain the total number of people in in the preset time period; and the number of people out at different collection time points can be added to obtain the total number of people out in the preset time period.

[0145] Here, the first people flow threshold and the second people flow threshold are pre-set, and the first people flow threshold and the second people flow threshold can be set according to actual needs. After obtaining the total number of people in and the total number of people out in the preset time period, it can be judged whether the total number of people in in the preset time period is greater than the set first people flow threshold, and / or whether the total number of people out in the preset time period is greater than the set second people flow threshold.

[0146] In the case of judging whether the total number of people in in the preset time period is greater than the set first people flow threshold, and judging whether the total number of people out in the preset time period is greater than the set second people flow threshold, if the total number of people in in the preset time period is greater than the set first people flow threshold, and / or if the total number of people out in the preset time period is greater than the set second people flow threshold, a people flow state alarm information is generated. The generated people flow state alarm information can be information in the form of text, voice, video, etc., for example, the generated people flow state alarm information can be "attention, the number of people in is large". In this case, the alarm event type of the people flow state alarm information is: point position crossing alarm.

[0147] Further, after triggering the generated people flow state alarm information, the detailed information of the people flow state alarm can be displayed, including but not limited to alarm point position (i.e. the name of the alarm monitoring device, etc.), alarm time, alarm event type, when the alarm event type is point position crossing alarm, the detailed information further includes the number of people in and the number of people out in the unit time, etc.

[0148] Here, when the monitoring point is one, the total number of people entering and the total number of people leaving in the monitoring area corresponding to the monitoring point in the preset time period are determined based on the number of people entering and the number of people leaving at different collection time points in the preset time period. When the total number of people entering in the preset time period is greater than the set first people flow threshold, and / or the total number of people leaving in the preset time period is greater than the set second people flow threshold, a people flow state alarm information is generated, realizing early warning of the number of people entering and the number of people leaving of the monitoring video, so as to guide the people flow based on the generated people flow state alarm information, avoiding the occurrence of safety accidents caused by a large number of people entering or a large number of people leaving in a short time.

[0149] In an optional implementation, when the monitoring point is one, the people flow state data of at least one monitoring device is determined based on the number of people monitoring data matching the monitoring event in the preset time period, including: determining the entering flow speed and the leaving flow speed in the monitoring area corresponding to the monitoring point based on the number of people entering and the number of people leaving at different collection time points in the preset time period.

[0150] Here, the entering flow speed and the leaving flow speed in the monitoring area corresponding to the monitoring point can also be determined based on the number of people entering and the number of people leaving at different collection time points in the preset time period.

[0151] In specific implementation, after obtaining the number of people entering and the number of people leaving at different collection time points in the preset time period, the multiple detection results can be classified and integrated according to the output time (collection time point) to obtain the number of people entering and the number of people leaving in a unit time (such as one minute), and then the entering flow speed and the leaving flow speed can be obtained.

[0152] For example, the output results within the output time of 08:10:00-08:11:00 (excluding 08:10:00 and including 08:11:00) can be classified and integrated, that is, the output results obtained at the output time of 08:10:03, 08:10:06, …, 08:10:57 and 08:11:00 are classified into one category, and the detection results in this category are integrated to obtain the number of people entering and the number of people leaving in 1 minute (unit time) between 08:10:00-08:11:00, that is, the entering flow speed (unit: person / min) and the leaving flow speed (unit: person / min) corresponding to 08:10 are obtained.

[0153] In the above method, the in-flow speed and the out-flow speed in the monitoring area corresponding to the monitoring point can be determined based on the in-flow quantity and the out-flow quantity at different collection time points in a preset time period, the in-flow speed and the out-flow speed are monitored, and the occurrence of a safety accident caused by a large in-flow speed or a large out-flow speed is avoided.

[0154] Referring to Figure 3 Fig. 3 shows a detailed interface diagram of a flow state alarm, which includes alarm details, cross-line event period statistics of the day, alarm details including alarm point, event type, alarm time, duration (cross-line event duration), in-flow peak value, out-flow peak value, and the like, and current cross-line event period statistics including cross-line event alarms from zero of the day to the current time of statistics. The diagram also includes a video screenshot, which displays the out-flow information (out-flow quantity and out-flow speed) and the in-flow information (in-flow quantity and in-flow speed) corresponding to the current time; and multiple alarm pictures are displayed below the video screenshot, wherein the number of alarm pictures is related to the duration of the alarm, for example, when the duration of the cross-line event is 17 minutes, one alarm picture can be extracted every minute as an alarm record, that is, 17 alarm pictures can be displayed below the video screenshot.

[0155] In specific implementation, the name, installation position, collected monitoring video, and other point information of the monitoring device, and the out-flow quantity and the in-flow quantity per unit time can be persistently stored in a search server (such as elasticsearch) for subsequent search and query.

[0156] In an optional implementation, when the monitoring point is multiple, the flow state data of at least one monitoring device is determined based on the number of people monitoring data matching the monitoring event in a preset time period, comprising:

[0157] Step one, for each monitoring point, the total in-flow quantity and the total out-flow quantity in the preset time period in the monitoring area corresponding to the monitoring point are determined based on the in-flow quantity and the out-flow quantity at different collection time points in the preset time period.

[0158] Step two, the personnel net stock in the target monitoring area is determined based on the historical number of people in the target monitoring area in the preset time period, and the total in-flow quantity and the total out-flow quantity in the preset time period corresponding to the multiple monitoring points respectively.

[0159] In the case that the flow state data of at least one monitoring device meets the alarm condition, the flow state alarm information is generated, including: in the case that the personnel net stock in the target monitoring area is greater than the set net stock threshold, the flow state alarm information is generated.

[0160] Here, considering that a site or place can be provided with multiple monitoring devices, the monitoring videos collected by the multiple monitoring devices can be analyzed for people flow to obtain people flow state data of the multiple monitoring videos. The monitoring videos collected by the multiple monitoring devices form a video group, that is, the video group can be analyzed for people flow to obtain people flow state data corresponding to the video group. In specific implementation, the cross-line event opening button corresponding to the video group set on the display interface can be used to open the cross-line analysis function of each monitoring video in the video group. At the same time, the specific information of the people flow total inventory hierarchical early warning can also be set on the display interface, for example, the first-level early warning number corresponding to the inventory trend, the second-level early warning number corresponding to the inventory warning, and the third-level early warning number corresponding to the inventory overheating.

[0161] In step one, for each monitoring point, after obtaining the in-flow quantity and out-flow quantity of different collection time points in the preset time period corresponding to the monitoring point, the in-flow quantity of different collection time points can be added to obtain the total in-flow quantity in the preset time period corresponding to the monitoring point, and the out-flow quantity of different collection time points can be added to obtain the total out-flow quantity in the preset time period corresponding to the monitoring point. Then the total in-flow quantity and the total out-flow quantity in the preset time period corresponding to each monitoring point can be obtained.

[0162] For example, the total in-flow quantity and the total out-flow quantity in the time period from 08:11:00 to 08:12:00 corresponding to each monitoring device can be obtained, and the time period between 08:11:00 and 08:12:00 is the preset time period.

[0163] Further, based on the historical number of the target monitoring area in the preset time period and the total in-flow quantity and the total out-flow quantity in the preset time period corresponding to the multiple monitoring points, the net inventory of personnel in the target monitoring area is determined. For example, for each monitoring video in the video group, the total in-flow quantity and the total out-flow quantity in the preset time period corresponding to the monitoring video are subtracted to obtain the people flow change quantity in the preset time period of the monitoring video, the people flow change quantities in the preset time period corresponding to each monitoring video are added to obtain the total people flow change quantity corresponding to the video group (i.e., the total people flow change quantity corresponding to the site or place corresponding to the video group), and the total people flow change quantity corresponding to the video group and the historical number of the target monitoring area in the preset time period are added to obtain the net inventory of personnel in the target monitoring area (i.e., the current number of people at the current time point corresponding to the site or place corresponding to the video group).

[0164] For example, the preset time period can be a time period between a time point of 08:11:00 to a time point of 08:12:00, the current number of people corresponding to the time point of 08:11:00 (i.e., the historical number of people of the target monitoring area in the preset time period) can be obtained, and the total number of people entering and the total number of people leaving in the video group corresponding to each monitoring video in the time period of 08:11:00 to 08:12:00 (the preset time period) can be obtained, and the net population of the target monitoring area at the time point of 08:12:00 is determined based on the historical number of people of the target monitoring area in the preset time period (i.e., the net population of the obtained time point of 08:11:00) and the total number of people entering and the total number of people leaving in the video group corresponding to each monitoring video in the time period of 08:11:00 to 08:12:00.

[0165] After obtaining the net population of the target monitoring area, the net population of the target monitoring area can be monitored, and when the net population of the target monitoring area is greater than the preset net population threshold, the flow state alarm information is generated. For example, the generated flow state alarm information can be "attention, the current time xx site population is more". In this case, the alarm event type of the flow state alarm information is: video group cross-line alarm.

[0166] In specific implementation, for the video group cross-line alarm, a plurality of alarm risks can be set, for example, the plurality of alarm risks include: population trend, population warning, and population overheating, different net population thresholds are set for different alarm risks, for example, the net population threshold corresponding to the population trend can be 100, the net population threshold corresponding to the population warning can be 200, and the net population threshold corresponding to the population overheating can be 500. Different flow state alarm information can be set for different alarm risks. For example, the flow state alarm information corresponding to the population trend can be: text format alarm information; the flow state alarm information corresponding to the population warning can be: voice format alarm information; and the flow state alarm information corresponding to the population overheating can be: video format alarm information.

[0167] Further, after triggering the generated flow state alarm information, the detailed information of the flow state alarm can be displayed, the detailed information includes but is not limited to alarm point (i.e., the name of the alarm monitoring device, etc.), alarm time, alarm event type, and when the alarm event type is the video group cross-line alarm, the detailed information can further include: the net population of the current time point.

[0168] Here, after determining the total number of people entering and leaving the monitoring area corresponding to each monitoring point within a preset time period, the net number of people in the target monitoring area can be determined based on the historical number of people in the target monitoring area within the preset time period, as well as the total number of people entering and leaving the target monitoring area within the preset time period corresponding to multiple monitoring points. If the net number of people in the target monitoring area is greater than the set net number threshold, a pedestrian flow status alarm message is generated, realizing an early warning of the net number of people in the target monitoring area. This allows for the guidance of people based on the generated pedestrian flow status alarm message when the net number of people is high, thus preventing safety accidents caused by a large number of people in the target monitoring area.

[0169] See Figure 4a The diagram shows an interface that displays detailed information about pedestrian flow status alarms. Figure 4a The map view displays detailed information about pedestrian flow status alarms; and see also... Figure 4b The diagram shows another interface that displays detailed information about pedestrian flow status alarms. Figure 4b The details of the pedestrian flow status alarms are displayed in a list mode. Figure 4b The list displayed includes over-density events and boundary-crossing events. Specifically, in the triggering... Figure 4a After the information about the line-crossing event is displayed, or after it is triggered Figure 4b After displaying the information about the line-crossing event, you can show... Figure 4c The alarm details are displayed in the image. Figure 4c The alarm details displayed include the group name (i.e., the name corresponding to the video group), event type, alarm time, duration, peak total number of people, and total number of people on the day.

[0170] Secondly, the alarm process for excessively dense events can be explained in detail.

[0171] In one optional implementation, when the monitored event is an overly dense event, the number of people monitored within a preset time period that matches the monitored event is obtained, including: counting the number of target objects at different collection time points within the preset time period.

[0172] Here, when monitoring markers corresponding to excessively dense events exist in the surveillance video, the target objects (humans) within the detection area corresponding to the monitoring markers can be detected based on the monitoring markers and the surveillance video, yielding the number of target objects within the detection area at each acquisition time point. When no monitoring markers corresponding to excessively dense events exist in the surveillance video, the entire monitoring screen is considered the detection area, and the surveillance video can be inspected to obtain the number of target objects within the detection area at each acquisition time point.

[0173] In a specific implementation, the trained deep learning algorithm for identifying the target object can be used to detect the detection region in the monitoring video, and the detection result can be output in real time. The detection result can be the number of people in the detection region at each collection time point in the monitoring video. The deep learning algorithm can output the detection result periodically, for example, the deep learning algorithm can output the detection result every second, or the deep learning algorithm can output the detection result every two seconds, and so on. For example, the detection result can be: the number of people in the detection region at 08:10:00 (collection time point) is 50; the number of people in the detection region at 08:10:01 is 54, and so on.

[0174] The number of target objects at different collection time points in a preset time period can be further counted, for example, the preset time period is from 08:10:00 to 08:11:00, and each second time point in the preset time period is regarded as a collection time point, that is, the number of target objects at 08:10:00 (collection time point 1), the number of target objects at 08:10:01 (collection time point 2), and so on.

[0175] In the above method, when the monitoring event is the over-crowding event, the number of target objects at different collection time points in a preset time period can be counted, which provides data support for determining the people flow state data corresponding to the over-crowding event.

[0176] In an optional implementation, when the monitoring point is one, the people flow state data of the at least one monitoring device is determined based on the number of people monitoring data matching the monitoring event in a preset time period, including: determining the average number of people in the monitoring area corresponding to the monitoring point in the preset time period based on the number of target objects at different collection time points in the preset time period.

[0177] When the people flow state data of the at least one monitoring device meets the alarm condition, the people flow state alarm information is generated, including: when it is determined that the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is greater than the set first number threshold, the people flow state alarm information is generated.

[0178] Here, the number of target objects at different collection time points in a preset time period can be averaged to obtain the average number of people in the monitoring area corresponding to the monitoring point in the preset time period. The average number of people in the preset time period is monitored, and when the average number of people is greater than the set first number threshold, the people flow state alarm information is generated. The length of the preset time period can be set as needed, for example, the length of the preset time period can be 5 seconds, 10 seconds, 60 seconds, 5 minutes, etc. The length of the preset time period corresponding to the cross-line event and the length of the preset time period corresponding to the over-crowding event can be the same or different.

[0179] For example, the number of target objects at different collection time points in the preset time period includes: the number of target objects at 08:10:01 is 50, the number of target objects at 08:10:02 is 53, the number of target objects at 08:10:03 is 52, the number of target objects at 08:10:04 is 51, and the number of target objects at 08:10:05 is 54. The average value of 5 detection results is obtained, which is 52, and it is determined that the average number of people in the monitoring area corresponding to the monitoring point is 52 from 08:10:01 to 08:10:05.

[0180] Further, the average number of people in the monitoring area corresponding to the monitoring point in the preset time period can be monitored, and when the average number of people is greater than the set first number threshold, a people flow state alarm information is generated. For example, the generated people flow state alarm information can be "attention, the current time xx area has more people". In this case, the alarm event type of the people flow state alarm information is: point position over density alarm.

[0181] Further, after triggering the generated people flow state alarm information, the detailed information of the people flow state alarm can be displayed, which includes but is not limited to alarm point position (i.e. the name of the alarm monitoring device and the like), alarm time, alarm event type, and when the alarm event type is point position over density alarm, the detailed information can further include: real-time number of people in the detection area at the current time point.

[0182] Referring to Figure 5 The figure includes alarm details, today's over density event period statistics, alarm details include alarm point position, event type, alarm time, over density duration, number of people peak, density peak, etc., and current over density event period statistics include over density event alarm from zero o'clock of the day to the current time of statistics. The figure also includes a video screenshot and multiple alarm pictures displayed below the video screenshot, wherein the number of alarm pictures is related to the duration of the over density event, for example, when the duration of the over density event is 17 minutes, one frame of alarm picture can be extracted every minute as an alarm record, that is, 17 frames of alarm pictures can be displayed below the video screenshot.

[0183] In specific implementation, the name of the monitoring device, the installation position, the collected monitoring video and other point position information, and the real-time number of people per minute of the monitoring device, the maximum value of the real-time number of people, the minimum value of the real-time number of people and other information can be persistently associated and stored in a search server (such as elasticsearch) for subsequent search query.

[0184] In the above method, when the monitoring point is one, the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is determined based on the number of target objects at different collection time points in the preset time period; and the crowd state alarm information is generated in the case that the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is greater than the set first number threshold, so as to realize the monitoring of the average number of people in the detection area of the monitoring video, so as to guide the crowd in the detection area based on the generated crowd state alarm information, and avoid the occurrence of safety accidents when the personnel in the detection area is relatively dense.

[0185] In an optional embodiment, when the monitoring point is multiple, the crowd state data of the at least one monitoring device is determined based on the number of people monitoring data matching the monitoring event in the preset time period, including:

[0186] Step one, for each monitoring point, the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is determined based on the number of target objects at different collection time points in the preset time period.

[0187] Step two, the total real-time number of people in the target monitoring area is determined based on the average number of people corresponding to each of the multiple monitoring points.

[0188] In the case that the crowd state data of the at least one monitoring device meets the alarm condition, the crowd state alarm information is generated, including: in the case that the total real-time number of people in the target monitoring area is greater than the set second number threshold, the crowd state alarm information is generated.

[0189] Here, for each monitoring point, the average number of people in the monitoring area corresponding to the monitoring point in the preset time period can be determined based on the number of target objects at different collection time points in the preset time period; and the total real-time number of people in the target monitoring area can be determined by adding the average number of people corresponding to each of the multiple monitoring points.

[0190] After determining the total real-time number of people in the target monitoring area, the total real-time number of people can be monitored, and the crowd state alarm information is generated when the total real-time number of people in the target monitoring area is greater than the set second number threshold.

[0191] In specific implementation, the opening button of the over-crowded event corresponding to the video group can be set on the display interface to open the crowd over-crowded analysis function of each monitoring video in the video group. At the same time, the specific information of real-time total number of people hierarchical early warning can also be set on the display interface, such as filling in the first level early warning number corresponding to general risk, the second level early warning number corresponding to larger risk, and the third level early warning number corresponding to major risk.

[0192] Here, when the monitoring device includes multiple, the monitoring videos collected by the multiple monitoring devices respectively constitute a video group. For each monitoring video collected by a monitoring device (i.e., for each monitoring video in the video group), a deep learning algorithm trained for identifying target objects can be used to detect the detection region indicated by the monitoring mark in the monitoring video, and real-time output the detection result. The detection result can be the number of target objects in the detection region at the time point of the monitoring video. Further, based on the detection result obtained periodically, the average number of people in the monitoring area corresponding to the monitoring point can be determined within a preset time period.

[0193] After obtaining the average number of people corresponding to each monitoring video in the video group, the average number of people corresponding to each monitoring video included in the video group can be added to determine the total real-time number of people in the target monitoring area. Further, the total real-time number of people in the target monitoring area can be monitored, and if the total real-time number of people in the target monitoring area is greater than a set second number threshold, a crowd state alarm information is generated. For example, the generated crowd state alarm information can be "Attention, the total number of people in the current scene is large". In this case, the alarm event type of the crowd state alarm information is: video group overcrowding alarm.

[0194] In specific implementation, for the video group overcrowding alarm, multiple alarm risks can be set, such as a first risk, a second risk, and a third risk. Different second number thresholds can be set for different alarm risks. For example, the second number threshold corresponding to the first risk can be 100, the second number threshold corresponding to the second risk can be 200, and the second number threshold corresponding to the third risk can be 500. Different crowd state alarm information can be set for different alarm risks. For example, the crowd state alarm information corresponding to the first risk can be text format alarm information; the crowd state alarm information corresponding to the second risk can be voice format alarm information; and the crowd state alarm information corresponding to the third risk can be video format alarm information.

[0195] Further, after the generated crowd state alarm information is triggered, detailed information of the crowd state alarm can be displayed. The detailed information includes but is not limited to alarm point (i.e., the name of the alarm monitoring device, etc.), alarm time, and alarm event type. When the alarm event type is video group overcrowding alarm, the detailed information can further include the total real-time number of people in the real scene.

[0196] Referring to Figure 6a , a detailed information of a crowd state alarm is displayed in a map mode; and referring to Figure 6a , another detailed information of a crowd state alarm is displayed in a map mode. Figure 4b , another detailed information of a crowd state alarm is displayed in a map mode.Figure 4b The details of the pedestrian flow status alarms are displayed in a list mode. Figure 4b The list displayed includes over-density events and boundary-crossing events. Specifically, in the triggering... Figure 6a After the information about the overly dense event is displayed, or after it is triggered Figure 4b After displaying information about overly sensitive events, you can show... Figure 6b The alarm details are displayed in the image. Figure 6b The alarm details displayed include group name (i.e., the name corresponding to the video group), event type, alarm time, duration, peak number of people, peak density, real-time total number of people for the day, and video source statistics.

[0197] Here, after determining the pedestrian flow status data in the real-world scene based on surveillance video collected by at least one monitoring device and pre-drawn monitoring markers matching the target locations in the video footage, a diagram illustrating the changes in pedestrian flow status data over time can be generated to provide a visual representation of the pedestrian flow status data for the day. Specifically, the diagram illustrating the changes in pedestrian flow status data over time includes a first diagram illustrating the changes in the total number of people in real time, which includes the relationship between peak and trough numbers of people over time; and / or a second diagram illustrating the changes in the total number of people over time, which includes the relationship between the total outflow of people, the total inflow of people over time, and the total number of people over time. The time intervals set for the first and second diagrams can be 5 minutes, 10 minutes, 30 minutes, 1 hour, etc.

[0198] In the above method, after determining the average number of people in the monitoring area corresponding to each monitoring point within a preset time period, the total real-time number of people in the target monitoring area can be determined based on the average number of people corresponding to multiple monitoring points respectively; and when the total real-time number of people in the target monitoring area is determined to be greater than the set second number threshold, a crowd flow status alarm is generated, thereby realizing an early warning of the total real-time number of people in the target monitoring area. In order to guide the people in the target monitoring area based on the generated crowd flow status alarm when the total real-time number of people is large, the occurrence of safety accidents caused by the large total real-time number of people in the target monitoring area can be avoided.

[0199] In one optional implementation, the method further includes: averaging the pedestrian flow status data at the same collection time point within multiple recent historical dates to obtain the predicted pedestrian flow status data corresponding to each collection time point; the predicted pedestrian flow status data corresponding to each collection time point constitutes the predicted pedestrian flow status data for future dates; wherein the predicted data is used to generate a pedestrian flow management plan.

[0200] Here, the plurality of historical periods can be set as needed, for example, the plurality of historical periods can be the flow state data in the recent 7 days (one historical period corresponds to one day), that is, at 00:00 on October 8, the flow state data from October 1 to October 7 (7 historical periods) can be obtained, the flow state data of the same collection time point in the recent 7 historical dates is averaged to obtain the prediction flow state data corresponding to each collection time point. The prediction flow state data corresponding to each collection time point constitutes the prediction data of the flow state data in the future date.

[0201] For example, the prediction data of the total number of people entering the flow in the future date (the next day), the prediction data of the total number of people leaving the flow in the future date (the next day), and the prediction data of the net stock of personnel in the future date (the next day) are generated.

[0202] Further, based on the prediction data of the flow state data in the future date, a flow dredging plan can be generated, for example, if it is known from the prediction data that the total real-time number is the largest at 15:00, the number of people entering the target monitoring area at 15:00 can be controlled.

[0203] In actual application scenarios, the method can be applied to scenarios such as shopping malls and halls. The following describes the overline events of a monitoring video and the overline events of a video group in a shopping mall, assuming that the shopping mall has two doors, a monitoring device can be set at each door position (monitoring point), that is, monitoring device one (monitoring device one set at monitoring point one) collects the monitoring video of door A, and monitoring device two (monitoring device two set at monitoring point two) collects the monitoring video of door B. The monitoring device one and the monitoring device two can monitor the pedestrians entering and leaving the door.

[0204] In specific implementation, the monitoring video one collected by the monitoring device one and the monitoring video two collected by the monitoring device two can be obtained. For the monitoring video one, the entry and exit boundary and the entry and exit direction are drawn on the video screenshot of the monitoring video one to determine the total number of people entering and the total number of people leaving the monitoring area corresponding to the monitoring point in the monitoring video one in a preset time period. Further, in the case that the total number of people entering in the preset time period is greater than the set first flow threshold, and / or the total number of people leaving in the preset time period is greater than the set second flow threshold, a flow state alarm information is generated. And for the monitoring video two, the entry and exit boundary and the entry and exit direction are set on the video screenshot of the monitoring video two to determine the total number of people entering and the total number of people leaving the monitoring area corresponding to the monitoring point in the monitoring video two in a preset time period. Further, in the case that the total number of people entering in the preset time period is greater than the set first flow threshold, and / or the total number of people leaving in the preset time period is greater than the set second flow threshold, a flow state alarm information is generated.

[0205] Meanwhile, the monitoring video one and the monitoring video two constitute a video group, and the video group can be analyzed to determine the people flow state data in the target monitoring area corresponding to the monitoring device one and the monitoring device two. In specific implementation, for the monitoring video one, the total in-flow quantity and the total out-flow quantity in the preset time period in the monitoring area corresponding to the monitoring point one are determined; for the monitoring video two, the total in-flow quantity and the total out-flow quantity in the preset time period in the monitoring area corresponding to the monitoring point two are determined. Then, based on the historical number of people in the target monitoring area in the preset time period and the total in-flow quantity and the total out-flow quantity in the preset time period corresponding to the plurality of monitoring points respectively, the net population in the target monitoring area is determined. That is, the net population in the mall is determined. In the case that the net population is greater than the set net population threshold, the people flow state alarm information is generated, so that after receiving the people flow state alarm information, the pedestrians in the mall can be regulated to avoid the occurrence of congestion events.

[0206] The following describes the over-crowding event of one monitoring video and the over-crowding event of a video group respectively with the hall as an example. It is assumed that the monitoring devices are respectively arranged at the four corners (four monitoring points) of the hall, that is, the four monitoring devices detect the four monitoring areas of the hall, and that the monitoring videos collected by each monitoring device in the four monitoring devices constitute a video group.

[0207] In specific implementation, for each monitoring video in the video group, based on the number of target objects at different collection time points in the preset time period, the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is determined, and when the average number of people corresponding to the monitoring video is greater than the set first number threshold, the people flow state alarm information corresponding to the monitoring video is generated. That is, the over-crowding event monitoring for each monitoring video in the video group is realized.

[0208] Meanwhile, the monitoring mark can be drawn on the video picture screenshot, that is, the area corresponding to the monitoring mark is the detection area; or the monitoring mark can not be drawn on the video picture screenshot, that is, the monitoring video does not have a corresponding reference surface mark, and in this case, the entire video picture is the detection area by default.

[0209] Meanwhile, the video group can be monitored for an over-crowding event to determine a total real-time number of people in a target monitoring area corresponding to the video group. In a specific implementation, for each monitoring video in the video group, an average number of people in a monitoring area corresponding to a monitoring point is determined within a preset time period; and based on the average number of people corresponding to the four monitoring points, a total real-time number of people in the target monitoring area is determined. That is, the total real-time number of people in the multiple detection areas in the hall is determined. In a case where the total real-time number of people in the target monitoring area is greater than a second number threshold set, a crowd state alarm information is generated, so that after the crowd state alarm information is received, the crowded area in the hall can be guided to avoid accidents caused by the crowded area.

[0210] Those skilled in the art can understand that, in the above method of the specific implementation, the writing order of each step does not mean a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0211] Based on the same concept, the embodiment of the present disclosure also provides a scene monitoring device, as shown in Figure 7 The architecture schematic diagram of the scene monitoring device provided by the embodiment of the present disclosure includes a first acquisition module 701, a detection module 702, a second acquisition module 703, and a determination module 704, and specifically:

[0212] The first acquisition module 701 is configured to acquire monitoring videos collected by monitoring devices arranged at at least one monitoring point.

[0213] The detection module 702 is configured to determine whether a monitoring event occurs in a monitoring area corresponding to the at least one monitoring point based on the monitoring videos.

[0214] The second acquisition module 703 is configured to acquire people monitoring data matching the monitoring event within a preset time period in a case where the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point.

[0215] The determination module 704 is configured to determine crowd state data of the at least one monitoring device based on the people monitoring data matching the monitoring event within the preset time period.

[0216] In a possible implementation, after the crowd state data of the at least one monitoring device is determined, the device further includes an alarm module 705 configured to:

[0217] In a case where the crowd state data of the at least one monitoring device meets an alarm condition, generate a crowd state alarm information.

[0218] In a possible implementation, in a case where the monitoring event is a cross-line event, the detection module 702, when determining whether a monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the monitoring video, is configured to:

[0219] determine, based on the monitoring video, whether there is a target object that crosses a target position matched with the pre-drawn in-out boundary in the monitoring area corresponding to the at least one monitoring point;

[0220] if yes, determine that a cross-line event occurs in the monitoring area corresponding to the at least one monitoring point.

[0221] In a possible implementation, in a case where the monitoring event is a cross-line event, the second acquisition module 703, when acquiring the people number monitoring data matched with the monitoring event in a preset time period, is configured to:

[0222] acquire the in-flow quantity and the out-flow quantity at different collection time points in the preset time period, where the in-flow quantity at the different collection time points refers to the number of people crossing the pre-drawn in-out boundary in a pre-drawn in direction at different collection time points, and the out-flow quantity at the different collection time points refers to the number of people crossing the pre-drawn in-out boundary in a pre-drawn out direction at different collection time points.

[0223] In a possible implementation, in a case where the monitoring point is one, the determination module 704, when determining the people flow state data of the at least one monitoring device based on the people number monitoring data matched with the monitoring event in a preset time period, is configured to:

[0224] determine, based on the in-flow quantity and the out-flow quantity at different collection time points in the preset time period, the total in-flow quantity and the total out-flow quantity in the monitoring area corresponding to the monitoring point in the preset time period;

[0225] The alarm module 705, when determining that the people flow state data of the at least one monitoring device meets an alarm condition, is configured to:

[0226] generate people flow state alarm information in a case where it is determined that the total in-flow quantity in the preset time period is greater than a set first flow threshold, and / or it is determined that the total out-flow quantity in the preset time period is greater than a set second flow threshold.

[0227] In a possible implementation, in a case where the monitoring point is one, the determination module 704, when determining the people flow state data of the at least one monitoring device based on the people number monitoring data matched with the monitoring event in a preset time period, is configured to:

[0228] Determine the in-flow speed and out-flow speed in the monitoring area corresponding to the monitoring point based on the in-flow quantity and out-flow quantity at different collection time points in the preset time period.

[0229] In a possible implementation, when the monitoring point is multiple, the determining module 704 is configured to:

[0230] For each monitoring point, determine the total in-flow quantity and total out-flow quantity in the preset time period in the monitoring area corresponding to the monitoring point based on the in-flow quantity and out-flow quantity at different collection time points in the preset time period.

[0231] Determine the personnel net stock in the target monitoring area based on the historical personnel quantity in the preset time period of the target monitoring area and the total in-flow quantity and total out-flow quantity in the preset time period of the monitoring point corresponding to the target monitoring area.

[0232] The alarm module 705 is configured to:

[0233] Generate the flow state alarm information when it is determined that the personnel net stock in the target monitoring area is greater than the set net stock threshold.

[0234] In a possible implementation, when the monitoring event is the over-density event, the detecting module 702 is configured to:

[0235] Determine whether the number of target objects in the monitoring area corresponding to the monitoring point exceeds the over-density threshold based on the monitoring video.

[0236] If yes, determine that the monitoring event occurs in the monitoring area corresponding to the monitoring point.

[0237] In a possible implementation, when the monitoring event is the over-density event, the second obtaining module 703 is configured to:

[0238] Count the number of target objects at different collection time points in the preset time period.

[0239] In a possible implementation, when the monitoring point is one, the determining module 704 is configured to:

[0240] determine the average number of people in the monitoring area corresponding to the monitoring point in the preset time period based on the number of target objects at different collection time points in the preset time period;

[0241] The alarm module 705 is configured to generate the people flow state alarm information when it is determined that the people flow state data of the at least one monitoring device meets the alarm condition.

[0242] generate the people flow state alarm information when it is determined that the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is greater than the first number threshold.

[0243] In a possible implementation, when the monitoring point is one, the determining module 704 is configured to:

[0244] determine the average number of people in the monitoring area corresponding to the monitoring point in the preset time period based on the number of target objects at different collection time points in the preset time period;

[0245] determine the total real-time number of people in the target monitoring area based on the average number of people corresponding to each of the monitoring points;

[0246] The alarm module 705 is configured to generate the people flow state alarm information when it is determined that the people flow state data of the at least one monitoring device meets the alarm condition.

[0247] generate the people flow state alarm information when it is determined that the total real-time number of people in the target monitoring area is greater than the second number threshold.

[0248] In a possible implementation, the apparatus further includes a pre-warning module 706, configured to:

[0249] average the people flow state data at the same collection time point in a plurality of historical dates to obtain predicted people flow state data corresponding to each collection time point;

[0250] The predicted people flow state data corresponding to each collection time point constitutes predicted data of the people flow state data in future dates; and the predicted data is used to generate a people flow diversion plan.

[0251] In some embodiments, the apparatus provided by the embodiments of the present disclosure has functions or contains templates that can be used to perform the methods described in the above method embodiments, and specific implementations can refer to the descriptions of the above method embodiments. For brevity, they will not be repeated here.

[0252] Based on the same technical concept, the embodiments of the present disclosure also provide an electronic device. Referring to Figure 8 As shown in FIG. 8, the electronic device 800 provided by the embodiments of the present disclosure includes a processor 801, a memory 802, and a bus 803. The memory 802 is used to store execution instructions, including an internal memory 8021 and an external memory 8022. The internal memory 8021 is also called an internal storage, and is used to temporarily store operation data in the processor 801 and exchange data with the external memory 8022 such as a hard disk. The processor 801 exchanges data with the external memory 8022 through the internal memory 8021. When the electronic device 800 is running, the processor 801 and the memory 802 communicate through the bus 803, so that the processor 801 executes the following instructions:

[0253] obtaining monitoring video collected by a monitoring device arranged at at least one monitoring point;

[0254] determining whether a monitoring event occurs in a monitoring area corresponding to the at least one monitoring point based on the monitoring video;

[0255] in a case where the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point, obtaining people number monitoring data matching the monitoring event in a preset time period;

[0256] determining people flow state data of the at least one monitoring device based on the people number monitoring data matching the monitoring event in the preset time period.

[0257] In addition, the embodiments of the present disclosure also provide a computer readable storage medium, which stores a computer program. When the computer program is run by a processor, the steps of the scene monitoring method described in the above method embodiments are executed.

[0258] The computer program product of the scene monitoring method provided by the embodiments of the present disclosure includes a computer readable storage medium storing program codes. The instructions included in the program codes can be used to execute the steps of the scene monitoring method described in the above method embodiments. For details, refer to the above method embodiments, which will not be repeated here.

[0259] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working process of the system and the device described above can refer to the corresponding process in the foregoing method embodiment, and will not be repeated here. In several embodiments provided in the present disclosure, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and there can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some communication interfaces, devices or units, and can be electrical, mechanical or other forms.

[0260] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., can be located in one place or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0261] In addition, the functional units in each embodiment of the present disclosure can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit.

[0262] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a non-volatile computer readable storage medium executable by a processor. Based on this understanding, the technical solutions of the present disclosure essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present disclosure. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0263] The above is only a specific embodiment of the present disclosure, but the protection scope of the present disclosure is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present disclosure, which should be covered within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims

1. A method of scene monitoring, characterized by, The method comprises: acquiring monitoring video collected by a monitoring device arranged at at least one monitoring point; determining whether a monitoring event occurs in a monitoring area corresponding to the at least one monitoring point based on the monitoring video; in a case where the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point, acquiring people number monitoring data matching the monitoring event in a preset time period; determining people flow state data of the at least one monitoring device based on the people number monitoring data matching the monitoring event in the preset time period; the monitoring event comprises a cross-line event or an over-density event; the people number monitoring data comprises at least one of the following: a number of in-flow and out-flow corresponding to the cross-line event, and a number of target objects corresponding to the over-density event.

2. The method of claim 1, wherein, after determining the people flow state data of the at least one monitoring device, the method further comprises: generating people flow state alarm information in a case where the people flow state data of the at least one monitoring device meets an alarm condition.

3. The method according to any of claims 1-2, characterized by, before determining whether the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the monitoring video, the method further comprises: generating a monitoring mark corresponding to the monitoring video and matching the monitoring event; wherein the monitoring mark comprises at least one of the following: an in-out boundary line, an in direction, an out direction, and a polygon structure; and / or for any video frame in the monitoring video, in response to a triggered human body labeling operation, determining human body bounding box information of a plurality of pedestrians located at different depth positions in the video frame, wherein the human body bounding box information comprises area information and depth information of the human body bounding box.

4. The method of claim 3, wherein, the generating the monitoring mark corresponding to the monitoring video and matching the monitoring event comprises: acquiring a video picture screenshot; wherein the video picture screenshot comprises a pre-drawn monitoring mark matching the monitoring event; determining position information of the monitoring mark in the video picture screenshot; based on the position information corresponding to the monitoring mark, generating the monitoring mark matching the monitoring event in a video picture of the monitoring video.

5. The method of claim 3, wherein, in a case where the monitoring event comprises an over-density event, the determining whether the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the monitoring video comprises: for a monitoring area corresponding to each monitoring point, determining a predicted area of the monitoring area corresponding to the monitoring mark based on human body bounding box information marked in the human body labeling operation; determining a people density corresponding to the monitoring area based on a detected number of people in the monitoring area and the predicted area; in a case where the people density is greater than a set value, determining that the monitoring event of over-density occurs in the monitoring area corresponding to the at least one monitoring point.

6. The method of claim 1, wherein, in a case where the monitoring event comprises a cross-line event, the determining whether the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the monitoring video comprises: determining, based on the monitoring video, whether there is a target object crossing a target position matching a pre-drawn in-out boundary line in the monitoring area corresponding to the at least one monitoring point; if there is, determining that the cross-line event occurs in the monitoring area corresponding to the at least one monitoring point.

7. The method of claim 2, wherein, In a case where the monitoring event includes a cross-line event and the monitoring point is one, the people flow state data of the at least one monitoring device is determined based on the people number monitoring data matching the monitoring event in a preset time period, including: The total in-flow quantity and the total out-flow quantity in the monitoring area corresponding to the monitoring point in the preset time period are determined based on the in-flow quantity and the out-flow quantity at different collection time points in the preset time period indicated by the people number monitoring data, and / or the in-flow speed and the out-flow speed in the monitoring area corresponding to the monitoring point are determined; wherein the in-flow quantity at different collection time points refers to the number of people crossing the pre-drawn in-out boundary in the in direction at different collection time points; the out-flow quantity at different collection time points refers to the number of people crossing the pre-drawn in-out boundary in the out direction at different collection time points; The people flow state data includes the total in-flow quantity and the out-flow quantity, and in a case where the people flow state data of the at least one monitoring device meets an alarm condition, people flow state alarm information is generated, including: In a case where the total in-flow quantity in the preset time period is greater than a set first flow threshold, and / or in a case where the total out-flow quantity in the preset time period is greater than a set second flow threshold, the people flow state alarm information is generated.

8. The method of claim 2, wherein, In a case where the monitoring event includes a cross-line event and the monitoring point is multiple, the people flow state data of the at least one monitoring device is determined based on the people number monitoring data matching the monitoring event in a preset time period, including: For each monitoring point, the total in-flow quantity and the total out-flow quantity in the monitoring area corresponding to the monitoring point in the preset time period are determined based on the in-flow quantity and the out-flow quantity at different collection time points in the preset time period indicated by the people number monitoring data; The personnel net stock in the target monitoring area is determined based on the historical people number in the target monitoring area in the preset time period and the total in-flow quantity and the total out-flow quantity in the preset time period corresponding to the multiple monitoring points respectively; In a case where the people flow state data of the at least one monitoring device meets an alarm condition, people flow state alarm information is generated, including: In a case where the personnel net stock in the target monitoring area is greater than a set net stock threshold, the people flow state alarm information is generated.

9. The method of claim 2, wherein, In a case where the monitoring event includes a cross-line event and the monitoring point is one, the people flow state data of the at least one monitoring device is determined based on the people number monitoring data matching the monitoring event in a preset time period, including: The average number of people in the monitoring area corresponding to the monitoring point in the preset time period is determined based on the number of target objects at different collection time points in the preset time period indicated by the people number monitoring data; In a case where the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is greater than a set first number threshold, the people flow state alarm information is generated.

10. The method of claim 2, wherein, In the case that the monitoring event includes an over-crowding event and the monitoring point is multiple, the people flow state data of the at least one monitoring device is determined based on the number of people monitoring data matching the monitoring event in a preset time period, including: For each monitoring point, the average number of people in the monitoring area corresponding to the monitoring point in the preset time period is determined based on the number of people monitoring data indicating the number of target objects at different collection time points in the preset time period; The total real-time number of people in the target monitoring area is determined based on the average number of people corresponding to each of the multiple monitoring points; In the case that the total real-time number of people in the target monitoring area is greater than a set second number threshold, a people flow state alarm information is generated.

11. The method according to any one of claims 1 to 10, characterized in that, The method further includes: The people flow state data at the same collection time point in the recent multiple historical dates is averaged to obtain the predicted people flow state data corresponding to each collection time point; The predicted people flow state data corresponding to each collection time point respectively constitutes the prediction data of the people flow state in the future dates; wherein the prediction data is used to generate a people flow dredging plan.

12. A scene monitoring apparatus characterized by comprising: It includes: The first acquisition module is used to acquire the monitoring video collected by the monitoring device arranged at the at least one monitoring point; The detection module is used to determine whether the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point based on the monitoring video; The second acquisition module is used to acquire the number of people monitoring data matching the monitoring event in a preset time period in the case that the monitoring event occurs in the monitoring area corresponding to the at least one monitoring point; The determination module is used to determine the people flow state data of the at least one monitoring device based on the number of people monitoring data matching the monitoring event in a preset time period.

13. An electronic device, comprising: It includes: The processor, the memory and the bus, the memory stores machine readable instructions executable by the processor, when the electronic device is running, the processor and the memory are communicated through the bus, the machine readable instructions are executed by the processor to execute the steps of the scene monitoring method in any one of claims 1 to 11.

14. A computer-readable storage medium, characterized in that, The computer program is stored on the computer readable storage medium, and the computer program is executed by the processor to execute the steps of the scene monitoring method in any one of claims 1 to 11.

Images