Small-range passenger flow prediction system and method based on space-time analysis

Abstract

The invention relates to the technical field of passenger flow management, in particular to a small-range passenger flow prediction system and method based on space-time analysis. The method comprises the following steps: acquiring a monitoring image in real time; analyzing the monitoring image, and determining monitoring area information and area activity information; determining the maximum aggregation amount according to the monitoring area information; determining a trigger event at the current moment according to the regional activity information; analyzing the trigger event, and determining event influence; according to the event influence and the maximum aggregation amount, predicting people flow change in a preset time period; and according to the people flow change, determining regional convergence information and sending the regional convergence information to a manager. The specific influence of the event on the small-range passenger flow, such as the number of attracted crowds and the residence time, is predicted, and a basis is provided for small-range passenger flow prediction; by predicting the passenger flow change in the preset time period, the management personnel can identify the passenger flow trend in advance, so that resource allocation and safety management can be better carried out.

Description

Technical Field

[0001] This application relates to the technical field of passenger flow management, and in particular, to a small-scale passenger flow prediction system and method based on spatio-temporal analysis. Background Art

[0002] With the continuous advancement of the urbanization process, the passenger flow in public places such as museums, scenic spots, shopping malls, exhibitions, and concerts is increasing day by day. The large-scale passenger flow gathering will not only bring great pressure to the managers, but also may bring potential safety hazards and increase the management difficulty.

[0003] When some hot events occur, the gathering of people often shows the characteristic of a rapid increase within a short period of time, and it is difficult to effectively predict and manage through general monitoring means. Therefore, how to effectively monitor, predict, and control the small-scale and sudden passenger flow density in real time has become an urgent problem to be solved. Summary of the Invention

[0004] This application provides a small-scale passenger flow prediction system and method based on spatio-temporal analysis to solve the above problems.

[0005] In a first aspect, this application provides a small-scale passenger flow prediction method based on spatio-temporal analysis, and the method includes: Obtain monitoring images in real time; analyze the monitoring images to determine the monitoring area information and area activity information; Determine the maximum gathering amount according to the monitoring area information; Determine the triggering event at the current moment according to the area activity information; analyze the triggering event to determine the event impact; Predict the change of the passenger flow within a preset time period according to the event impact and the maximum gathering amount; Determine the area convergence information according to the change of the passenger flow and send it to the management personnel.

[0006] Through this solution, real-time collection of surveillance images can ensure the freshness and accuracy of data. By analyzing the surveillance images, the passenger flow dynamics in the surveillance area can be grasped in real time, including the passenger flow density, movement direction, residence time, etc., thus providing real-time data support for passenger flow prediction. By determining the maximum gathering volume, it can provide a reference for safety management to ensure that the passenger flow in public places does not exceed the safety carrying capacity and avoid crowded and trampling accidents. Identifying the triggering events at the current moment helps the management staff analyze the reasons for the changes in passenger flow, and thus take more targeted measures. By analyzing the impact of events, the specific impact of events on passenger flow can be predicted, such as the number of people attracted, residence time, etc., providing a basis for passenger flow prediction. By predicting the changes in passenger flow within a preset time period, it helps the management staff identify the passenger flow trend in advance, and thus better allocate resources and conduct safety management. By determining the regional convergence information, it helps the management staff identify which areas will be crowded, and thus take measures in advance, such as increasing staff, setting up temporary channels, etc., to ensure the safety of the crowd.

[0007] Optionally, the analysis of the surveillance images to determine the surveillance area information includes: Obtain the GIS map of the surveillance area, and the GIS map contains the geographical coordinates of all surveillance devices; Obtain the installation information of the surveillance devices, and determine the installation parameters and installation perspectives according to the installation information; Based on the geographical coordinates, analyze the installation parameters and installation perspectives of each surveillance device to determine the coverage area of each surveillance device; Perform spatial overlay of all surveillance coverage areas with the GIS map to determine the surveillance area information.

[0008] Through this solution, the GIS map provides the geographical coordinates of the monitoring area, which helps to determine the coverage range of the monitoring devices. The installation information of the monitoring devices includes parameters such as the installation location, angle, and height of the devices, which helps to analyze the coverage range and installation perspective of the monitoring devices. Based on the installation parameters, the effective monitoring range of the monitoring devices can be calculated, including the horizontal and vertical perspectives, which helps to divide and analyze the monitoring area. Determining the installation perspective helps to avoid mutual occlusion of the monitoring devices and ensure the coherence and integrity of the monitoring area. Through the geographical coordinates and installation parameters, the coverage area of each monitoring device can be accurately calculated, including the monitoring range and blind area of the device. Analyzing the coverage area helps to determine the overall layout of the monitoring area, identify the monitoring blind areas, and provide a basis for the adjustment and optimization of the monitoring devices. Through the spatial overlay of the GIS map and the coverage range of the monitoring devices, a complete monitoring area model can be constructed. The spatial overlay helps to identify the hot spots in the monitoring area and provide more accurate data support for passenger flow prediction and management. Determining the monitoring area information helps the management personnel to better identify the structure and function of the monitoring area and provide basic data for passenger flow prediction and management.

[0009] Optionally, determining the maximum aggregation quantity according to the monitoring area information includes: Determining the effective bearing area according to the GIS map; Calculating the maximum number of people that can be accommodated per unit area according to the preset safety specifications; Determining the area of each monitoring area according to the monitoring area information; Determining the maximum aggregation quantity according to the effective bearing area, the area of the region, and the maximum number of people that can be accommodated.

[0010] Through this solution, through GIS map analysis, the effective bearing area of the monitoring area can be accurately determined, which helps to conduct passenger flow management and safety assessment. Referring to the preset safety specifications, calculating the maximum number of people that can be accommodated per unit area provides a basis for the safety management of the monitoring area and ensures that people can evacuate safely in case of an emergency. Through the analysis of the monitoring area information, the monitoring area of each monitoring area can be accurately calculated, providing data support for passenger flow prediction and area management. Combining the effective bearing area, the area of the region, and the maximum number of people that can be accommodated per unit area, the maximum aggregation quantity of the monitoring area can be calculated, providing an important reference for passenger flow management and safety control. By accurately calculating the maximum aggregation quantity, overcrowding of people can be prevented in advance, reducing safety risks and ensuring that people can evacuate safely in case of an emergency.

[0011] Optionally, determining the maximum aggregation quantity according to the effective bearing area, the area of the region, and the maximum number of people that can be accommodated includes: Determining the actual bearing area of each monitoring area according to the effective bearing area and the area of the region; Obtain a real-time heat map; analyze the real-time heat map to determine the change in the confluence of people; Determine the convergence peak according to the change in the confluence of people; Determine the maximum aggregation amount according to the maximum capacity and the convergence peak.

[0012] Through this solution, by analyzing the GIS map, the actual carrying area of the monitored area can be accurately calculated, which helps the management personnel to identify the available space in the monitored area, so as to facilitate the planning of the crowd flow route and the safety evacuation plan. The real-time heat map provides an intuitive view of the crowd distribution in the monitored area, which helps to monitor and manage the passenger flow dynamics in real time, and at the same time helps to adjust the management strategy in a timely manner to cope with emergencies. By analyzing the real-time heat map, the hot spots and flow trends of crowd gathering can be identified, which helps to predict the crowd flow pattern and identify potential congestion and safety hazards in advance. Analyzing the change in the confluence of people can identify the flow pattern and speed of people in the monitored area, which helps to optimize the layout of the monitored area and the crowd flow control measures, and improve the crowd flow efficiency. Determining the convergence peak of people can predict the peak time of passenger flow, which helps to take preventive measures in advance to ensure the safety and orderly flow of people during the peak time. Combining the maximum capacity and the convergence peak, the maximum aggregation amount of the monitored area can be calculated, providing a safety upper limit for passenger flow management, which helps to evaluate the safety risk of the monitored area, formulate an emergency plan, and ensure the safe evacuation of people in case of emergency.

[0013] Optionally, the analyzing the trigger event to determine the event impact includes: Analyze the trigger event to determine the event type and trigger time; Obtain the historical area information according to the trigger time; Analyze the historical area information to determine the historical pedestrian flow corresponding to the trigger time; Determine the event impact according to the historical pedestrian flow and the event type.

[0014] Through this solution, real-time data is collected through channels such as monitoring devices, sensors, social media, and news sources to identify trigger events that can cause changes in passenger flow. This helps management personnel promptly identify events affecting passenger flow and provides a basis for passenger flow prediction and management. Classifying the trigger events helps identify the nature and impact of the events. Determining the event time provides a time reference for obtaining and analyzing historical area information. Based on the trigger time, historical area information is extracted to provide a data basis for event impact analysis. Analyzing the historical passenger flow corresponding to the trigger time to identify the changes in passenger flow when the event occurs helps evaluate the degree of impact of the event on passenger flow. Providing historical data support for passenger flow prediction and management helps identify patterns and trends in passenger flow changes. Based on the analysis results of historical area information, determining the historical passenger flow corresponding to the trigger time helps identify the degree of impact of the event on passenger flow and provides a reference for passenger flow prediction and management. Combining the historical passenger flow and the event type to evaluate the event impact, including the analysis of the degree of attraction to passenger flow, duration, crowd behavior patterns, etc., provides the event impact evaluation results for passenger flow prediction and management and helps formulate corresponding management strategies and preventive measures.

[0015] Optionally, determining the event impact according to the historical passenger flow and the event type includes: Analyze the monitoring image to determine the passenger flow data at the current moment; Based on the passenger flow data, determine whether there is a targeted customer group for the trigger event at the current moment; If so, determine the historical passenger flow change according to the historical area information; Based on the historical passenger flow change, predict the stop flow of the targeted customer group at the trigger time; Determine the personnel appeal according to the event type; Based on the historical passenger flow, the stop flow, and the personnel appeal, determine the event impact.

[0016] Through this solution, by analyzing surveillance images, real-time pedestrian flow data in the surveillance area can be obtained, including the number of people, crowd density, flow direction, etc., which helps to monitor the changes in pedestrian flow in real time and provide instant information for passenger flow management. By analyzing the pedestrian flow data, targeted customer groups related to trigger events can be identified, such as fans of star events, shoppers in promotional activities, etc., which helps to predict the impact of the event on the pedestrian flow. Through historical area information, the impact of past similar events on the pedestrian flow can be identified, providing a reference for the prediction of the current event. By analyzing historical pedestrian flow changes, the number of people and duration of stay of the targeted customer group in the surveillance area at the trigger time can be predicted, providing a basis for resource allocation and pedestrian flow management. According to the event type, the attractiveness of the event to the crowd can be evaluated, thereby predicting the passenger flow scale brought by the event, which helps to prepare corresponding resources and management measures in advance. Combining historical pedestrian flow, predicted stop flow, and the appeal of personnel of the event type, the impact of the event on the pedestrian flow can be comprehensively evaluated, including passenger flow scale, crowd gathering risk, etc., providing support for formulating management strategies and safety plans.

[0017] Optionally, predicting the stop flow of the targeted customer group at the trigger time according to the historical pedestrian flow changes includes: Based on the event type, analyze the targeted customer group to determine the stickiness between the targeted customer group and the trigger event; According to the historical pedestrian flow changes, determine the average stay duration; According to the stickiness and the average stay duration, predict the stop flow of the targeted customer group at the trigger time.

[0018] Through this solution, by analyzing the event type, the characteristics and preferences of the targeted customer group can be identified, which helps to predict the response and participation degree of the targeted customer group to the trigger event. Stickiness analysis helps to evaluate the loyalty and participation degree of the targeted customer group to the trigger event, thereby predicting the average stay duration of the targeted customer group and the impact of the event. By analyzing historical pedestrian flow data, the average stay duration can be determined, which helps to predict the stay time of the targeted customer group when the trigger event occurs. Combining stickiness and average stay duration, the stop flow of the targeted customer group at the time of the event can be predicted, thus providing a basis for management.

[0019] Optionally, determining the regional convergence information according to the pedestrian flow changes includes: According to the pedestrian flow changes, determine whether there is a possibility of space congestion currently; If there is, according to the pedestrian flow changes, determine the congested area, and determine the congested area and the pedestrian flow changes as the regional convergence information.

[0020] Through this solution, real-time collection of pedestrian flow data in the monitored area helps managers analyze the changes in pedestrian flow in real time and provides basic data for passenger flow management. By analyzing the real-time pedestrian flow data, the flow patterns, crowded areas, and potential safety risks of the crowd can be identified, providing a basis for management decisions. The application of the crowding detection algorithm can automatically identify the possibility of space crowding, improve the efficiency and accuracy of detection, and reduce the errors of human judgment. The identification of crowded areas helps managers quickly locate problem areas and take evacuation, guidance, or other safety measures in a timely manner to avoid excessive crowd gathering. Combining the crowded area and the pedestrian flow change data to form regional convergence information provides a comprehensive information view for managers, helping to manage passenger flow more effectively.

[0021] Optionally, the method further includes: After sending to the manager, continuously detect the pedestrian flow image of the crowded area; Analyze the pedestrian flow image to determine the reduction of pedestrian flow within a preset time period; According to the event type and the reduction of pedestrian flow, determine whether the pedestrian flow returns to the expected state.

[0022] Through this solution, through continuous monitoring of the pedestrian flow image, managers can analyze the changes in the pedestrian flow in the crowded area in real time, timely grasp the situation of crowd movement, and provide a basis for adjusting management strategies. By analyzing the pedestrian flow image, the trend of pedestrian flow changes within a preset time period can be accurately identified, including the speed and pattern of the number of people decreasing, providing data support for predicting the return of pedestrian flow to the expected state. Combining the event type and the reduction of pedestrian flow, it can be evaluated whether the pedestrian flow returns to the expected state as expected, which helps managers judge whether it is necessary to continue to take control measures or adjust management strategies. According to whether the pedestrian flow returns to the expected state, managers can timely adjust management strategies, such as optimizing the pedestrian flow path, increasing evacuation channels, guiding the crowd movement, etc., to ensure crowd safety and public order.

[0023] In a second aspect, the present application provides a small-scale passenger flow prediction system based on spatio-temporal analysis. The system includes: An image analysis module, configured to obtain a monitoring image in real time; analyze the monitoring image to determine the monitoring area information and the regional activity information; A regional information analysis module, configured to determine the maximum aggregation amount according to the monitoring area information; An event analysis module, configured to determine the triggering event at the current moment according to the regional activity information; analyze the triggering event to determine the event impact; A pedestrian flow prediction module, configured to predict the change in pedestrian flow within a preset time period according to the event impact and the maximum aggregation amount; An information sending module, configured to determine area aggregation information according to the change in the flow of people and send it to the management personnel.

[0024] Optionally, when the image analysis module analyzes the monitoring image to determine the monitoring area information, it is used for: Obtain a GIS map of the monitoring area, where the GIS map contains the geographical coordinates of all monitoring devices; Obtain the installation information of the monitoring device, and determine the installation parameters and installation perspective according to the installation information; Based on the geographical coordinates, analyze the installation parameters and installation perspective of each monitoring device to determine the coverage area of each monitoring device; Perform a spatial overlay of all monitoring coverage areas with the GIS map to determine the monitoring area information.

[0025] Optionally, when the area information analysis module determines the maximum aggregation amount according to the monitoring area information, it is used for: Determine the effective bearing area according to the GIS map; Calculate the maximum number of people that can be accommodated per unit area according to the preset safety specifications; Determine the area of each monitoring area according to the monitoring area information; Determine the maximum aggregation amount according to the effective bearing area, the area of the monitoring area, and the maximum number of people that can be accommodated.

[0026] Optionally, when the area information analysis module determines the maximum aggregation amount according to the effective bearing area, the area of the monitoring area, and the maximum number of people that can be accommodated, it is used for: Determine the actual bearing area of each monitoring area according to the effective bearing area and the area of the monitoring area; Obtain a real-time heat map; analyze the real-time heat map to determine the change in the flow of people; Determine the convergence peak according to the change in the flow of people; Determine the maximum aggregation amount according to the maximum number of people that can be accommodated and the convergence peak.

[0027] Optionally, when the event analysis module analyzes the trigger event to determine the event impact, it is used for: Analyze the trigger event to determine the event type and trigger time; Obtain the historical area information according to the trigger time; Analyze the historical area information to determine the historical flow of people corresponding to the trigger time; Determine the event impact according to the historical flow of people and the event type.

[0028] Optionally, when the event analysis module determines the event impact based on the historical pedestrian flow and the event type, it is used for: Analyze the monitoring image to determine the pedestrian flow data at the current moment; Based on the pedestrian flow data, determine whether there is a targeted customer group for the trigger event at the current moment; If so, determine the historical pedestrian flow change according to the historical area information; Based on the historical pedestrian flow change, predict the stop flow of the targeted customer group at the trigger time; Determine the personnel appeal according to the event type; Determine the event impact based on the historical pedestrian flow, the stop flow, and the personnel appeal.

[0029] Optionally, when the event analysis module predicts the stop flow of the targeted customer group at the trigger time according to the historical pedestrian flow change, it is used for: Based on the event type, analyze the targeted customer group to determine the stickiness between the targeted customer group and the trigger event; Determine the average stay duration according to the historical pedestrian flow change; Predict the stop flow of the targeted customer group at the trigger time based on the stickiness and the average stay duration.

[0030] Optionally, when the information sending module determines the area convergence information according to the pedestrian flow change, it is used for: Determine whether there is a possibility of space congestion currently according to the pedestrian flow change; If so, determine the congested area according to the pedestrian flow change, and determine the congested area and the pedestrian flow change as the area convergence information.

[0031] Optionally, the small-scale passenger flow prediction system based on spatio-temporal analysis further includes a state determination module, which is used for: Continuously detect the pedestrian flow image of the congested area after sending it to the management personnel; Analyze the pedestrian flow image to determine the reduction of the pedestrian flow within a preset time period; Determine whether the pedestrian flow returns to the expected state according to the event type and the reduction of the pedestrian flow. Description of the Drawings

[0032] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other accompanying drawings can be obtained based on these drawings without creative efforts.

[0033] Figure 1 A schematic diagram of an application scenario provided by an embodiment of the present application; Figure 2 A flowchart of a small-scale passenger flow prediction method based on spatio-temporal analysis provided by an embodiment of the present application; Figure 3 A schematic diagram of the structure of a small-scale passenger flow prediction system based on spatio-temporal analysis provided by an embodiment of the present application. Detailed implementation manners

[0034] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts fall within the scope of protection of the present application.

[0035] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after unless otherwise specified.

[0036] The following will further describe the embodiments of the present application in detail with reference to the accompanying drawings of the specification.

[0037] When some hot events occur, the gathering of people often shows the characteristic of a rapid increase within a short period of time, making it difficult to effectively predict and manage through general monitoring means. Therefore, how to monitor, predict, and control the small-scale and sudden passenger flow density situation in real time has become an urgent problem to be solved.

[0038] Based on this, the present application provides a small-scale passenger flow prediction system and method based on spatio-temporal analysis, which can obtain monitoring images in real time; analyze the monitoring images to determine the monitoring area information and regional activity information; determine the maximum aggregation amount according to the monitoring area information; determine the triggering event at the current moment according to the regional activity information; analyze the triggering event to determine the event impact; predict the change of the passenger flow within a preset time period according to the event impact and the maximum aggregation amount; and determine the regional convergence information according to the change of the passenger flow and send it to the management personnel. Through this solution, collecting monitoring images in real time can ensure the freshness and accuracy of the data. By analyzing the monitoring images, the dynamic state of the passenger flow within the monitoring area can be grasped in real time, including the passenger flow density, moving direction, staying time, etc., so as to provide real-time data support for passenger flow prediction. By determining the maximum aggregation amount, it can provide a reference for safety management to ensure that the passenger flow in public places does not exceed the safety carrying capacity and avoid the occurrence of crowding and trampling accidents. Identifying the triggering event at the current moment helps the management personnel analyze the reasons for the change of the passenger flow, so as to take more targeted measures. By analyzing the event impact, the specific impact of the event on the passenger flow can be predicted, such as the number of attracted people, staying time, etc., providing a basis for passenger flow prediction. By predicting the change of the passenger flow within a preset time period, it helps the management personnel identify the passenger flow trend in advance, so as to better allocate resources and manage safety. By determining the regional convergence information, it helps the management personnel identify which areas will be crowded, so as to take measures in advance, such as increasing the number of staff, setting up temporary channels, etc., to ensure the safety of the crowd.

[0039] Figure 1 This is a schematic diagram of an application scenario provided by the present application. When predicting the small-scale passenger flow volume, the method provided by the present application is applied. Specifically, the method provided by the present application is applied to any server, and the server interacts with the monitoring device. Collecting the monitoring images uploaded by the monitoring device in real time can ensure the freshness and accuracy of the data. By analyzing the monitoring images, the dynamic state of the passenger flow within the monitoring area can be grasped in real time, including the passenger flow density, moving direction, staying time, etc., so as to provide real-time data support for passenger flow prediction. By determining the maximum aggregation amount, it can provide a reference for safety management to ensure that the passenger flow in public places does not exceed the safety carrying capacity and avoid the occurrence of crowding and trampling accidents. Identifying the triggering event at the current moment helps the management personnel analyze the reasons for the change of the passenger flow, so as to take more targeted measures. By analyzing the event impact, the specific impact of the event on the passenger flow can be predicted, such as the number of attracted people, staying time, etc., providing a basis for passenger flow prediction. By predicting the change of the passenger flow within a preset time period, it helps the management personnel identify the passenger flow trend in advance, so as to better allocate resources and manage safety. By determining the regional convergence information, it helps the management personnel identify which areas will be crowded, so as to take measures in advance, such as increasing the number of staff, setting up temporary channels, etc., to ensure the safety of the crowd.

[0040] For the specific implementation method, reference can be made to the following embodiments.

[0041] Figure 2 The figure below is a flowchart of a small-scale passenger flow prediction method based on spatio-temporal analysis provided by an embodiment of the present application. The method of this embodiment can be applied to the server in the above scenarios. As Figure 2 shown, the method includes: S201. Obtain monitoring images in real time; analyze the monitoring images to determine the monitoring area information and regional activity information; The monitoring images can be real-time images captured by monitoring devices.

[0042] The monitoring area information can be geographical information such as the location, boundary, and structure of the monitoring area, as well as parameters such as the installation location and coverage range of the monitoring devices.

[0043] The regional activity information can be various activities such as crowd movement, aggregation, and dispersion occurring in the monitoring area.

[0044] Specifically, deploy monitoring devices in public places, and use the monitoring devices to capture monitoring images in real time. Through network transmission, the monitoring images are transmitted to the data processing center. Use computer vision algorithms to identify individuals such as pedestrians and vehicles in the monitoring images. Use image processing technologies such as object detection, crowd counting, and tracking algorithms to extract crowd information from the monitoring images. Analyze the behavior patterns of the crowd, such as staying, aggregating, and dispersing. Use GIS maps to determine the monitoring area information. Analyze the movement trajectories of individuals in the monitoring area, identify the flow patterns of the crowd such as entering, leaving, staying, aggregating, and dispersing, and determine the regional activity information.

[0045] S202. Determine the maximum aggregation amount according to the monitoring area information; The maximum aggregation amount can be the maximum number of people that can be accommodated in the monitoring area.

[0046] Specifically, use GIS map data to determine the effective bearing area within the monitoring area information. According to safety specifications, building standards, fire regulations, etc., determine the maximum number of people that can be accommodated per unit area. Multiply the effective bearing area by the maximum number of people that can be accommodated per unit area to obtain the maximum aggregation amount.

[0047] S203. Determine the trigger event at the current moment according to the regional activity information; analyze the trigger event to determine the event impact; The current moment can be the specific time point for passenger flow prediction.

[0048] The trigger event can be a hot event such as a star event, a promotional activity, a program performance, or an accident that causes changes in the flow of people.

[0049] The event impact can be the specific impact of a triggering event on the flow of people.

[0050] Specifically, real-time obtain regional activity information through monitoring devices, and use event recognition algorithms to identify hot events such as star events, promotional activities, performances, etc. that will cause changes in the passenger flow. According to the recognition results, determine the type and triggering time of the triggering event. According to the type and triggering time of the triggering event, extract information on similar events from historical data. Analyze the impacts of historical events on the attraction of the flow of people, duration, crowd behavior patterns, etc. Combine the regional activity information to evaluate the event impact of the triggering event at the current moment.

[0051] S204. Predict the change in the flow of people within a preset time period according to the event impact and the maximum aggregation volume; The preset time period can be a preset future time period for passenger flow prediction, which is stored in the server in advance and called when in use.

[0052] The change in the flow of people can be the change trend of the number, density, distribution, etc. of the flow of people in the monitored area within the preset time period.

[0053] Specifically, in a museum or scenic area, due to the rapid surge of people in a short period of time, which will quickly dissipate after a short period of maintenance, it is difficult to manage and predict. At the same time, distinguish the secondary aggregation situation of the crowd. For example, under the trigger of a hot event (star event, product promotion, program performance, accident, etc.), the aggregation density within 5 minutes can reach 3-5 times the normal value in a small area, resulting in a secondary outbreak situation. Then, in the case of the occurrence of a small-area aggregation, it is possible to use the prediction of multiple small-area aggregations, behavior trajectories, the possibility of outbreaks, etc. to analyze the nature, scale, expected appeal, etc. of the triggering event, and predict the change in the flow of people brought about by the triggering event. Analyze the maximum aggregation volume of the triggering event. Set the preset time period for prediction, that is, the time range for which the change in the flow of people needs to be predicted. According to the event type and the maximum aggregation volume, select an appropriate prediction model. According to the event impact and the maximum aggregation volume, set the parameters of the prediction model. Use the prediction model to predict the change in the flow of people within the preset time period according to the event impact and the maximum aggregation volume.

[0054] S205. Determine the regional aggregation information according to the change in the flow of people and send it to the management personnel.

[0055] The regional aggregation information can be the potential crowded area and the flow density information determined according to the prediction result of the change in the flow of people.

[0056] The management personnel can be the management personnel responsible for the safety management and resource allocation of public places.

[0057] Specifically, a monitoring device is used to collect real-time changes in the flow of people in the monitored area, such as crowd density, flow direction, and flow speed. The real-time changes in the flow of people are analyzed. Based on the analysis results of the flow of people changes, regional convergence information is determined. According to the regional convergence information, the regional convergence information is sent to the management terminal.

[0058] Through this solution, collecting monitoring images in real time can ensure the freshness and accuracy of data. By analyzing the monitoring images, the dynamic of customer flow in the monitored area can be grasped in real time, including crowd density, moving direction, staying time, etc., so as to provide real-time data support for customer flow prediction. By determining the maximum aggregation volume, it can provide a reference for safety management to ensure that the flow of people in public places does not exceed the safety carrying capacity, and avoid crowded and trampling accidents. Identifying the triggering event at the current moment helps the management personnel analyze the reasons for the changes in customer flow, so as to take more targeted measures. By analyzing the impact of the event, the specific impact of the event on customer flow can be predicted, such as the number of people attracted and the staying time, etc., providing a basis for customer flow prediction. By predicting the changes in the flow of people within a preset time period, it helps the management personnel identify the customer flow trend in advance, so as to better allocate resources and manage safety. By determining the regional convergence information, it helps the management personnel identify which areas will be crowded, so as to take measures in advance, such as increasing staff and setting up temporary passages, etc., to ensure the safety of the crowd.

[0059] In some embodiments, a GIS map of the monitored area is obtained. The GIS map contains the geographical coordinates of all monitoring devices; the installation information of the monitoring devices is obtained, and the installation parameters and installation perspectives are determined according to the installation information; based on the geographical coordinates, the installation parameters and installation perspectives of each monitoring device are analyzed to determine the coverage area of each monitoring device; all the monitoring coverage areas are spatially superimposed with the GIS map to determine the monitoring area information.

[0060] The monitored area can be areas such as museum exhibition halls, scenic area visitor centers, shopping mall shopping areas that need to be monitored.

[0061] The GIS map can be a digital map containing spatial data such as geographical location information, land use, building structure, and transportation network.

[0062] The monitoring devices can be devices such as cameras and sensors used to monitor partial areas.

[0063] The geographical coordinates can be used to determine the precise position of the monitoring device and the specific position of the monitored area.

[0064] The installation information can be parameters such as the installation position, angle, and height of the monitoring device.

[0065] The installation perspective can be the perspective range that the monitoring device can observe.

[0066] The covered area can be the area range that the monitoring device can cover.

[0067] The monitoring covered area can be the sum of the areas covered by all monitoring devices.

[0068] Specifically, extract the GIS map of the monitoring area from the GIS database. Integrate the geographical coordinate information of all monitoring devices into the GIS map. Collect installation information such as the model, installation location, installation angle, and installation height of the monitoring devices from the device installation records. Determine the installation parameters of the monitoring devices according to the collected installation information. Determine the installation perspective of the monitoring devices according to the installation parameters. Determine the specific location of each monitoring device in the monitoring area according to the geographical coordinates of the monitoring devices in the GIS map. Calculate the monitoring range of the monitoring devices according to the installation parameters such as the installation angle, installation height, and viewing angle range of each monitoring device. Analyze the installation perspective of each monitoring device to identify the blind spots and dead angles of the monitoring devices. Combine the geographical coordinates, installation parameters, and installation perspective, and use spatial analysis tools to calculate the covered area of each monitoring device. Digitalize the covered area of each monitoring device. Use GIS software to perform a spatial overlay operation to merge the digitized all monitoring covered areas with the GIS map, so as to determine the monitoring area information.

[0069] Through this solution, the GIS map provides the geographical coordinates of the monitoring area, which helps to determine the coverage range of the monitoring devices. The installation information of the monitoring devices includes parameters such as the installation location, angle, and height of the devices, which helps to analyze the coverage range and installation perspective of the monitoring devices. Through the installation parameters, the effective monitoring range of the monitoring devices, including the horizontal viewing angle and vertical viewing angle, can be calculated, which helps to divide and analyze the monitoring area. The determination of the installation perspective helps to avoid the mutual occlusion of the monitoring devices and ensure the coherence and integrity of the monitoring area. Through the geographical coordinates and installation parameters, the covered area of each monitoring device, including the monitoring range and blind spots of the device, can be accurately calculated. The analysis of the covered area helps to determine the overall layout of the monitoring area, identify the monitoring blind spots, and provide a basis for the adjustment and optimization of the monitoring devices. Through the spatial overlay of the GIS map and the covered area of the monitoring devices, a complete monitoring area model can be constructed. The spatial overlay helps to identify the hot spots in the monitoring area and provide more accurate data support for passenger flow prediction and management. The determination of the monitoring area information helps the management personnel to better identify the structure and function of the monitoring area and provide basic data for passenger flow prediction and management.

[0070] In some embodiments, determine the effective bearing area according to the GIS map; calculate the maximum number of people that can be accommodated per unit area according to the preset safety specifications; determine the area of each monitoring area according to the monitoring area information; determine the maximum aggregation amount according to the effective bearing area, the area of the area, and the maximum number of people that can be accommodated.

[0071] The effective load area can be the spatial area within the entire region that is actually available for accommodating people.

[0072] The preset safety specifications can be safety standards or regulations formulated for public places, which are formulated by government agencies, industry associations, or safety experts. They are pre-stored in the server and called when in use. The unit area can be the basic unit area used for calculating the maximum number of people that can be accommodated.

[0073] The maximum number of people that can be accommodated can be the maximum number of people that can be accommodated within the monitored area within the scope permitted by the safety specifications.

[0074] The area of the region can be the actual area corresponding to the monitored area.

[0075] Specifically, using the GIS map, determine the geographical information such as the building layout, public areas, and emergency exits of the monitored area. Analyze the spatial structure of the monitored area, identify the areas that can accommodate people, and exclude areas such as walls, columns, and equipment that are not accessible. According to the analysis results of the GIS map, calculate the effective load area of the monitored area. Collect the safety specifications applicable to the monitored area, analyze the requirements for population density in the safety specifications, and determine the maximum number of people that can be accommodated per unit area. Use the GIS map and monitoring equipment to analyze the information of the monitored area. According to the information of the monitored area, calculate the area of each monitored area. Multiply the effective load area of each monitored area by the maximum number of people that can be accommodated per unit area to calculate the maximum aggregation volume of each monitored area.

[0076] Through this solution, through GIS map analysis, the effective load area of the monitored area can be accurately determined, which helps with passenger flow management and safety assessment. Referring to the preset safety specifications, calculate the maximum number of people that can be accommodated per unit area, providing a basis for the safety management of the monitored area to ensure the safe evacuation of people in case of an emergency. Through the analysis of the information of the monitored area, the monitored area of each monitored area can be accurately calculated, providing data support for passenger flow prediction and area management. Combining the effective load area, the area of the region, and the maximum number of people that can be accommodated per unit area, the maximum aggregation volume of the monitored area can be calculated, providing an important reference for passenger flow management and safety control. By accurately calculating the maximum aggregation volume, overcrowding can be prevented in advance, reducing safety risks and ensuring the safe evacuation of people in case of an emergency.

[0077] In some embodiments, according to the effective load area and the area of the region, determine the actual load area of each monitored area; obtain the real-time heat map; analyze the real-time heat map to determine the change in the convergence of people; according to the change in the convergence of people, determine the convergence peak value; according to the maximum number of people that can be accommodated and the convergence peak value, determine the maximum aggregation volume.

[0078] The actual load-bearing area can be the spatial area within the monitored area that is actually available for accommodating people.

[0079] A real-time heat map can be a data visualization technique used to display the distribution of people in the monitored area.

[0080] The changes in the convergence of people can be the dynamic changes such as the aggregation, dispersion, and flow speed of the crowd flow in the monitored area.

[0081] The convergence peak value can be the number of people corresponding to the peak period of people converging.

[0082] Specifically, analyze the GIS map, identify the space available for accommodating people, and exclude areas such as non-accessible buildings and obstacles. Based on the identification results of the GIS map, calculate the effective load-bearing area of the monitored area. According to the information of the monitored area, determine the area of each monitored area. Exclude areas such as non-accessible walls, columns, and equipment from the area to determine the actual load-bearing area. Obtain the real-time heat map data of the monitored area through monitoring equipment. Analyze the real-time heat map to identify the dense areas and flow trends of people flow. According to the identification results of the real-time heat map, determine the changes such as the aggregation, dispersion, and flow speed of the convergence of people. Analyze the changes in the convergence of people to determine the peak time of people converging, that is, the most intensive moment of people flow within a certain time period. According to the preset safety specifications and the effective load-bearing area of the monitored area, calculate the maximum number of people that can be accommodated in the monitored area. Combine the maximum number of people that can be accommodated and the convergence peak value to calculate the maximum aggregation volume of the monitored area.

[0083] Through this solution, by analyzing the GIS map, accurately calculate the actual load-bearing area of the monitored area, which helps managers identify the available space in the monitored area for planning crowd flow routes and safety evacuation plans. The real-time heat map provides an intuitive view of the distribution of people in the monitored area, which helps to monitor and manage the passenger flow dynamics in real time, and at the same time helps to adjust management strategies in a timely manner to cope with emergencies. By analyzing the real-time heat map, hot spots and flow trends of crowd aggregation can be identified, which helps to predict crowd flow patterns and identify potential congestion and safety hazards in advance. Analyzing the changes in the convergence of people can identify the flow patterns and speeds of people in the monitored area, which helps to optimize the layout of the monitored area and crowd flow control measures and improve the efficiency of crowd flow. Determining the peak value of people converging can predict the peak period of passenger flow, which helps to take preventive measures in advance to ensure the safety and orderly flow of people during the peak period. Combining the maximum number of people that can be accommodated and the convergence peak value can calculate the maximum aggregation volume of the monitored area, providing a safety upper limit for passenger flow management, which helps to evaluate the safety risks of the monitored area, formulate emergency plans, and ensure the safe evacuation of people in case of emergencies.

[0084] In some embodiments, the triggering event is analyzed to determine the event type and the triggering time; according to the triggering time, historical area information is obtained; the historical area information is analyzed to determine the historical pedestrian flow corresponding to the triggering time; according to the historical pedestrian flow and the event type, the event impact is determined.

[0085] The event type can be the classification of hot events such as star activities, promotional activities, program performances, accidents, etc. that cause changes in passenger flow.

[0086] The triggering time can be the specific occurrence time of the triggering event.

[0087] The historical area information can be historical information such as the historical pedestrian flow in the monitored area and the crowd behavior pattern.

[0088] The historical pedestrian flow can be the pedestrian flow data in the monitored area during the time period corresponding to the triggering event.

[0089] Specifically, real-time data is collected through channels such as monitoring devices, sensors, social media, news sources, etc. to identify hot events such as star activities, promotional activities, program performances, accidents, etc. that cause changes in passenger flow. According to the identification result, the triggering event is classified to determine the time type of the triggering event. According to the identified triggering event, the triggering time of the triggering event, that is, the start time and the end time, is determined. According to the triggering time, the time stamp is matched with the records in the historical database to extract the historical area information corresponding to the triggering time. According to the triggering time, the historical area information is matched with the time series. The historical pedestrian flow corresponding to the triggering time is extracted from the historical area information. According to the historical pedestrian flow and the event type, the event impact is evaluated.

[0090] Through this solution, real-time data is collected through channels such as monitoring devices, sensors, social media, news sources, etc. to identify triggering events that cause changes in passenger flow, which helps managers promptly identify events that affect passenger flow and provides a basis for passenger flow prediction and management. Classifying the triggering events helps identify the nature and impact of the events. Determining the event time provides a time reference for obtaining and analyzing historical area information. According to the triggering time, extracting historical area information provides a data basis for event impact analysis. Analyzing the historical pedestrian flow corresponding to the triggering time to identify the changes in pedestrian flow when the event occurs helps evaluate the impact degree of the event on pedestrian flow. Providing historical data support for passenger flow prediction and management helps identify the patterns and trends of pedestrian flow changes. According to the analysis result of the historical area information, determining the historical pedestrian flow corresponding to the triggering time helps identify the impact degree of the event on pedestrian flow and provides a reference for passenger flow prediction and management. Combining the historical pedestrian flow and the event type to evaluate the event impact, including the analysis of the attraction degree, duration, crowd behavior pattern, etc. of the event on pedestrian flow, provides an event impact evaluation result for passenger flow prediction and management, which helps formulate corresponding management strategies and preventive measures.

[0091] In some embodiments, the monitored images are analyzed to determine the pedestrian flow data at the current moment; based on the pedestrian flow data, it is determined whether there is a targeted customer group that triggers an event at the current moment; if so, according to the historical area information, the historical change in pedestrian flow is determined; based on the historical change in pedestrian flow, the stop flow of the targeted customer group at the trigger time is predicted; according to the event type, the appeal of the personnel is determined; based on the historical pedestrian flow, stop flow and appeal of the personnel, the impact of the event is determined.

[0092] The pedestrian flow data can be the pedestrian flow data in the monitored area during the current time period.

[0093] The targeted customer group can be a customer group with shopping and other needs related to events such as promotional activities.

[0094] The historical change in pedestrian flow can be the change in pedestrian flow in the monitored area in the historical time series.

[0095] The stop flow can be the number of people and the duration of stay in the monitored area during the trigger time of the triggered event.

[0096] The appeal of the personnel can be the attraction of the triggered event to the crowd.

[0097] Specifically, in the case of the generation of a small-scale aggregation where the aggregation density can reach 3-5 times the normal value within 5 minutes under the trigger of a hot event (such as a star event, product promotion, program performance, accident, etc.), by predicting multiple small-scale aggregations, behavior trajectories, the possibility of outbreaks, etc., the nature, scale, expected appeal, etc. of the triggered event are analyzed to predict the change in pedestrian flow brought about by the triggered event. The collected monitored images are preprocessed such as denoising, adjusting brightness and contrast, and cropping. Computer vision algorithms are used to detect the monitored images to determine the pedestrian flow data at the current moment. Based on the pedestrian flow data, the behavior patterns of the crowd are analyzed, and classification algorithms are used to classify the crowd. According to the characteristics and behavior analysis results, it is judged whether there is a targeted customer group for the triggered event at the current moment. If there is a targeted customer group for the triggered event, the historical pedestrian flow is queried according to the triggered event and the trigger time. The historical pedestrian flow is analyzed to identify the pattern of change in pedestrian flow when past similar events occurred, that is, the historical change in pedestrian flow. Based on the historical change in pedestrian flow, a trend model is established. Using the trend model and the historical change in pedestrian flow, the stop flow of the targeted customer group at the trigger time is predicted. The event type is analyzed to identify the nature, scale, expected influence, etc. of the event type. According to the event type, the appeal of the event to the personnel is evaluated. Combining the historical pedestrian flow, the predicted stop flow and the appeal of the event type to the personnel, the impact of the event is thus determined.

[0098] Through this solution, by analyzing surveillance images, the pedestrian flow data in the surveillance area can be obtained in real time, including the number of people, crowd density, flow direction, etc., which helps to monitor the changes in pedestrian flow in real time and provide immediate information for passenger flow management. By analyzing the pedestrian flow data, it is possible to identify the targeted customer groups related to the triggering event, such as fans of a star event, shoppers during a promotional event, etc., which helps to predict the impact of the event on the pedestrian flow. Through the historical area information, it is possible to identify the impact of past similar events on the pedestrian flow and provide a reference for the prediction of the current event. By analyzing the historical changes in pedestrian flow, it is possible to predict the number of people and the duration of stay of the targeted customer groups in the surveillance area at the triggering time, providing a basis for resource allocation and pedestrian flow management. According to the event type, it is possible to evaluate the attractiveness of the event to the crowd, thereby predicting the passenger flow scale brought by the event, which helps to prepare the corresponding resources and management measures in advance. Combining the historical pedestrian flow, the predicted stop flow, and the appeal of the event type to people, it is possible to comprehensively evaluate the impact of the event on the pedestrian flow, including the passenger flow scale, crowd gathering risk, etc., providing support for formulating management strategies and safety plans.

[0099] In some embodiments, based on the event type, analyze the targeted customer group to determine the stickiness between the targeted customer group and the triggering event; according to the historical changes in pedestrian flow, determine the average stay duration; based on the stickiness and the average stay duration, predict the stop flow of the targeted customer group at the triggering time.

[0100] The stickiness can be the loyalty and participation of the targeted customer group to a certain triggering event.

[0101] The average stay duration can be the average length of time that the crowd stays in the surveillance area.

[0102] Specifically, analyze the event type to identify the nature, scale, expected influence, etc. of the event type. Analyze the targeted customer group to identify the characteristics, behavior patterns, preferences, etc. of the targeted customer group. Based on the event type and the characteristics of the targeted customer group, analyze the stickiness between the targeted customer group and the triggering event. Collect the historical changes in pedestrian flow such as the pedestrian flow and stay duration when past similar events occurred. Analyze the historical changes in pedestrian flow to determine the average stay duration. Based on the historical changes in pedestrian flow, establish a stop flow prediction model. Use the historical pedestrian flow data to train the prediction model. Use the trained prediction model to predict the stop flow of the targeted customer group at the triggering time based on the stickiness and the average stay duration.

[0103] Through this solution, by analyzing the event types and identifying the characteristics and preferences of the targeted customer group, it helps to predict the response and participation level of the targeted customer group to the triggering event. Stickiness analysis helps to evaluate the loyalty and engagement of the targeted customer group with the triggering event, thereby predicting the average stay duration and event impact of the targeted customer group. By analyzing historical footfall data, the average stay duration can be determined, which helps to predict the stay time of the targeted customer group when the triggering event occurs. Combining stickiness and average stay duration, the stop flow of the targeted customer group during the event can be predicted, providing a basis for management.

[0104] In some embodiments, according to the change in footfall, it is determined whether there is a possibility of space congestion currently; if so, according to the change in footfall, the congested area is determined, and the congested area and the change in footfall are determined as area convergence information.

[0105] The possibility of space congestion can be a state where the number or density of people in the monitored area reaches or approaches the safe accommodation limit, a potential situation that may trigger congestion and safety problems.

[0106] The congested area can be a specific area in the monitored area where the number or density of people exceeds the normal level, resulting in restricted movement of the crowd, decreased comfort, or potential safety risks.

[0107] Specifically, monitoring devices are used to collect real-time footfall data of the monitored area. The real-time footfall data is analyzed for the number of people, crowd density, flow direction, speed, etc. to determine the change in footfall at the current moment. A congestion detection algorithm is applied to analyze the change in footfall to determine whether there is a possibility of space congestion. If a possibility of space congestion is identified, the change in footfall is further analyzed to identify the specific congested area. The identified congested area and the change in footfall are combined and determined as area convergence information.

[0108] Through this solution, real-time collection of footfall data in the monitored area helps managers to analyze the change in footfall in real time, providing basic data for passenger flow management. By analyzing the real-time footfall data, the flow pattern of the crowd, the dense area, and potential safety risks can be identified, providing a basis for management decisions. The application of the congestion detection algorithm can automatically identify the possibility of space congestion, improve the efficiency and accuracy of detection, and reduce the error of human judgment. The identification of the congested area helps managers quickly locate the problem area and take evacuation, guidance, or other safety measures in a timely manner to avoid excessive crowd gathering. Combining the congested area and the footfall change data to form area convergence information provides managers with a comprehensive information view, helping to manage passenger flow more effectively.

[0109] In some embodiments, after sending to the management personnel, continuously detect the pedestrian flow images in the crowded area; analyze the pedestrian flow images to determine the reduction of pedestrian flow within a preset time period; and determine whether the pedestrian flow returns to the expected state according to the event type and the reduction of pedestrian flow.

[0110] The pedestrian flow images can be image data showing the changes in pedestrian flow captured by monitoring devices.

[0111] The reduction of pedestrian flow can be the situation where the number of pedestrians in the monitored area decreases within a preset time period.

[0112] The expected state can be the ideal number of pedestrians and flow pattern in the monitored area under the condition of crowded pedestrian flow.

[0113] Specifically, use the monitoring device to continuously capture the pedestrian flow images in the crowded area. Transmit the captured pedestrian flow images to the management personnel terminal. Use computer vision technology to analyze the pedestrian flow images, such as crowd counting, crowd density estimation, and detection of the direction of crowd flow. Detect abnormal situations in the pedestrian flow images, such as sudden gatherings of people and abnormal flow patterns. Evaluate the crowd density through the analysis of the pedestrian flow images. Analyze the direction and speed of the crowd flow to identify the pedestrian flow data. Perform time series analysis on the pedestrian flow data within a preset time period, and evaluate the reduction of pedestrian flow within the preset time period according to the results of the time series analysis. Analyze the event type, and identify the nature, scale, expected influence, etc. of the event type. Set the expected state according to the event type and the reduction of pedestrian flow. Compare the actual reduction of pedestrian flow with the expected state, and judge whether the pedestrian flow returns to the expected state.

[0114] Through this solution, by continuously monitoring the pedestrian flow images, the management personnel can analyze the changes in the pedestrian flow in the crowded area in real time, timely grasp the situation of crowd flow, and provide a basis for adjusting the management strategy. By analyzing the pedestrian flow images, the trend of changes in the pedestrian flow within a preset time period can be accurately identified, including the speed and pattern of the decrease in the number of people, providing data support for predicting the return of the pedestrian flow to the expected state. Combining the event type and the reduction of pedestrian flow, it can be evaluated whether the pedestrian flow returns to the expected state as expected, which helps the management personnel judge whether it is necessary to continue to take control measures or adjust the management strategy. According to whether the pedestrian flow returns to the expected state, the management personnel can timely adjust the management strategy, such as optimizing the pedestrian flow path, increasing evacuation channels, guiding the crowd flow, etc., to ensure the safety of the crowd and public order.

[0115] Figure 3 FIG. is a schematic structural diagram of a small-scale passenger flow prediction system based on spatio-temporal analysis provided in an embodiment of the present application. As Figure 3 shown, the small-scale passenger flow prediction system 300 based on spatio-temporal analysis in this embodiment includes: an image analysis module 301, a regional information analysis module 302, an event analysis module 303, a passenger flow prediction module 304, and an information sending module 305.

[0116] An image analysis module 301, configured to obtain surveillance images in real time; analyze the surveillance images to determine surveillance area information and area activity information; An area information analysis module 302, configured to determine the maximum aggregation amount according to the surveillance area information; An event analysis module 303, configured to determine the triggered event at the current moment according to the area activity information; analyze the triggered event to determine the event impact; A crowd flow prediction module 304, configured to predict the change of crowd flow within a preset time period according to the event impact and the maximum aggregation amount; An information sending module 305, configured to determine area convergence information according to the change of crowd flow and send it to the management personnel.

[0117] Optionally, when the image analysis module 301 analyzes the surveillance images to determine the surveillance area information, it is used for: Obtain the GIS map of the surveillance area, where the GIS map contains the geographical coordinates of all surveillance devices; Obtain the installation information of the surveillance devices, and determine the installation parameters and installation perspectives according to the installation information; Based on the geographical coordinates, analyze the installation parameters and installation perspectives of each surveillance device to determine the coverage area of each surveillance device; Perform a spatial overlay of all surveillance coverage areas with the GIS map to determine the surveillance area information.

[0118] Optionally, when the area information analysis module 302 determines the maximum aggregation amount according to the surveillance area information, it is used for: Determine the effective bearing area according to the GIS map; Calculate the maximum number of people that can be accommodated per unit area according to the preset safety specifications; Determine the area of each surveillance area according to the surveillance area information; Determine the maximum aggregation amount according to the effective bearing area, the area of the area, and the maximum number of people that can be accommodated.

[0119] Optionally, when the area information analysis module 302 determines the maximum aggregation amount according to the effective bearing area, the area of the area, and the maximum number of people that can be accommodated, it is used for: Determine the actual bearing area of each surveillance area according to the effective bearing area and the area of the area; Obtain the real-time heat map; analyze the real-time heat map to determine the change of personnel convergence; Determine the convergence peak according to the change of personnel convergence; Determine the maximum aggregation amount according to the maximum number of people that can be accommodated and the convergence peak.

[0120] Optionally, when the event analysis module 303 analyzes the triggered event to determine the event impact, it is used for: Analyze the triggering event to determine the event type and triggering time; Obtain historical area information according to the triggering time; Analyze the historical area information to determine the historical pedestrian flow corresponding to the triggering time; Determine the event impact according to the historical pedestrian flow and event type.

[0121] Optionally, when the event analysis module 303 determines the event impact according to the historical pedestrian flow and event type, it is used for: Analyze the monitored images to determine the pedestrian flow data at the current moment; Determine whether there is a targeted customer group triggering the event at the current moment according to the pedestrian flow data; If so, determine the historical pedestrian flow change according to the historical area information; Predict the stopping flow of the targeted customer group at the triggering time according to the historical pedestrian flow change; Determine the personnel appeal according to the event type; Determine the event impact according to the historical pedestrian flow, stopping flow and personnel appeal.

[0122] Optionally, when the event analysis module 303 predicts the stopping flow of the targeted customer group at the triggering time according to the historical pedestrian flow change, it is used for: Analyze the targeted customer group based on the event type to determine the stickiness between the targeted customer group and the triggering event; Determine the average stay duration according to the historical pedestrian flow change; Predict the stopping flow of the targeted customer group at the triggering time according to the stickiness and average stay duration.

[0123] Optionally, when the information sending module 305 determines the area convergence information according to the pedestrian flow change, it is used for: Determine whether there is a possibility of spatial congestion currently according to the pedestrian flow change; If so, determine the congested area according to the pedestrian flow change, and determine the congested area and pedestrian flow change as the area convergence information.

[0124] Optionally, the small-scale passenger flow prediction system based on spatio-temporal analysis further includes a status determination module 306, which is used for: Continuously detect the pedestrian flow images of the congested area after sending to the management personnel; Analyze the pedestrian flow images to determine the reduction of pedestrian flow within a preset time period; Determine whether the pedestrian flow returns to the expected state according to the event type and the reduction of pedestrian flow.

[0125] The system of this embodiment can be used to execute the method of any of the above embodiments, and its implementation principle and technical effects are similar, which will not be elaborated here.

Claims

1. A small-scale passenger flow prediction method based on spatio-temporal analysis, characterized in that, Including: Obtain monitoring images in real time; Analyze the monitoring images to determine monitoring area information and area activity information; Determine the maximum aggregation amount according to the monitoring area information; Determine the triggering event at the current moment according to the area activity information; Analyze the triggering event to determine the event impact; Predict the change in the flow of people within a preset time period according to the event impact and the maximum aggregation amount; Determine area convergence information according to the change in the flow of people and send it to the management personnel.

2. The method according to claim 1, characterized in that, The analyzing the monitoring images to determine the monitoring area information includes: Obtain the GIS map of the monitoring area, and the GIS map contains the geographical coordinates of all monitoring devices; Obtain the installation information of the monitoring devices, and determine the installation parameters and installation perspectives according to the installation information; Based on the geographical coordinates, analyze the installation parameters and installation perspectives of each monitoring device to determine the coverage area of each monitoring device; Perform spatial overlay on all monitoring coverage areas and the GIS map to determine the monitoring area information.

3. The method according to claim 2, wherein The determining the maximum aggregation amount according to the monitoring area information includes: Determine the effective bearing area according to the GIS map; Calculate the maximum number of people that can be accommodated per unit area according to the preset safety specifications; Determine the area of each monitoring area according to the monitoring area information; Determine the maximum aggregation amount according to the effective bearing area, the area of the monitoring area, and the maximum number of people that can be accommodated.

4. The method according to claim 3, wherein The determining the maximum aggregation amount according to the effective bearing area, the area of the monitoring area, and the maximum number of people that can be accommodated includes: Determine the actual bearing area of each monitoring area according to the effective bearing area and the area of the monitoring area; Obtain the real-time heat map; analyze the real-time heat map to determine the change in the convergence of people; Determine the convergence peak according to the change in the convergence of people; Determine the maximum aggregation amount according to the maximum number of people that can be accommodated and the convergence peak.

5. The method according to claim 1, characterized in that, The analyzing the triggering event to determine the event impact includes: Analyze the triggering event to determine the event type and the triggering time; Obtain the historical area information according to the triggering time; Analyze the historical area information to determine the historical flow of people corresponding to the triggering time; Determine the event impact according to the historical flow of people and the event type.

6. The method according to claim 5, characterized in that, The determining the event impact according to the historical flow of people and the event type includes: Analyze the monitoring images to determine the flow of people data at the current moment; Determine whether there is a targeted customer group for the triggering event at the current moment according to the flow of people data; If so, determine the historical change in the flow of people according to the historical area information; Predict the stop flow of the targeted customer group at the triggering time according to the historical change in the flow of people; Determine the appeal of people according to the event type; Determine the event impact according to the historical flow of people, the stop flow, and the appeal of people.

7. The method according to claim 6, wherein The predicting the stop flow of the targeted customer group at the triggering time according to the historical change in the flow of people includes: Based on the event type, analyze the targeted customer group to determine the stickiness between the targeted customer group and the triggering event; Determine the average stay duration according to the historical change in the flow of people; Predict the stop flow of the targeted customer group at the trigger time according to the stickiness and the average residence time.

8. The method according to claim 1, characterized in that, Determine the regional convergence information according to the change in the flow of people, including: Determine whether there is a possibility of spatial congestion according to the change in the flow of people; If so, determine the congested area according to the change in the flow of people, and determine the congested area and the change in the flow of people as the regional convergence information.

9. The method according to any one of claims 7-8, characterized in that, The method further includes: After sending to the management personnel, continuously detect the flow of people image in the congested area; Analyze the flow of people image to determine the reduction of the flow of people within a preset period; Determine whether the flow of people returns to the expected state according to the event type and the reduction of the flow of people.

10. A small-scale passenger flow prediction system based on spatio-temporal analysis, characterized in that, Including: An image analysis module for real-time acquisition of monitoring images; Analyze the monitoring image to determine the monitoring area information and the regional activity information; A regional information analysis module for determining the maximum aggregation amount according to the monitoring area information; An event analysis module for determining the trigger event at the current moment according to the regional activity information; Analyze the trigger event to determine the event impact; A flow of people prediction module for predicting the change in the flow of people within a preset period according to the event impact and the maximum aggregation amount; An information sending module for determining the regional convergence information according to the change in the flow of people and sending it to the management personnel.

Images