People flow measurement system, people flow measurement device, people flow measurement method, and people flow measurement program

Abstract

A moving body (200) is provided with a camera (210) and acquires video data (20) in a measurement region. A measurement unit (110) measures the number of people in each individual region of a plurality of individual regions as an actual measurement result (21) by using the video data (20). A prediction unit (120) predicts the number of people in each individual region of the plurality of individual regions as a prediction result (22) by using the actual measurement result (21) and a congestion prediction model (82). On the basis of the prediction result (22) and the actual measurement result (21), an analysis unit (160) identifies, as a specific region, an individual region having the lowest prediction accuracy among the plurality of individual regions, and estimates, as a specific people flow, a people flow having a correlation with the congestion in the specific region. A route determination unit (170) determines a movement route (31) for the moving body (200) to acquire video data of the specific people flow.

Description

People flow measurement system, people flow measurement device, people flow measurement method, and people flow measurement program

[0001] The present disclosure relates to a people flow measurement system, a people flow measurement device, a people flow measurement method, and a people flow measurement program.

[0002] In recent years, congestion status display systems have been developed to grasp the actual level of congestion in facilities with high traffic volumes, such as airports, train stations, shopping malls, theme parks, and event venues.

[0003] Patent Document 1 discloses a congestion monitoring system that uses a surveillance camera to estimate the degree of congestion in an entire facility and visualizes the estimated degree of congestion.

[0004] International Publication No. 2017 / 122258

[0005] In Patent Document 1, it is assumed that the surveillance cameras are installed so that they can capture the points that people always pass through when they move around, without any blind spots. Therefore, there is a problem in that it is costly to install additional cameras in places where cameras are not installed.

[0006] The present disclosure aims to reduce the number of sensors and installation costs without reducing prediction accuracy by appropriately measuring the flow of people in a measurement area using a moving object.

[0007] The people flow measurement system according to the present disclosure comprises: a mobile body that moves along a movement path in a measurement area having a plurality of individual areas, the mobile body being equipped with a camera and acquiring video data of the measurement area using the camera; a measurement unit that uses the video data to measure the number of people in each of the plurality of individual areas as an actual measurement result; a prediction unit that predicts the number of people in each of the plurality of individual areas as a prediction result using the actual measurement result and a congestion prediction model that includes a route model that models the route of people flow in the measurement area and a structural model that represents the structure of the measurement area; an analysis unit that identifies the individual area among the plurality of individual areas with the lowest prediction accuracy as a specific area based on the prediction result and the actual measurement result, and estimates a people flow that is correlated with congestion in the specific area as a specific people flow; and a route determination unit that determines the movement path that the mobile body will take to acquire video data of the specific people flow.

[0008] The people flow measurement system according to the present disclosure measures people flow in a measurement area using moving objects moving along appropriate paths, thereby achieving the effect of reducing the number of sensors and reducing installation costs without reducing prediction accuracy.

[0009] FIG. 1 is a diagram showing an example of the hardware configuration of a people flow measurement system according to embodiment 1. FIG. 2 is a diagram showing an example of the functional configuration of a people flow measurement system according to embodiment 1. FIG. 3 is a flow diagram showing prediction processing and result determination correction processing in people flow measurement processing according to embodiment 1. FIG. 4 is a schematic diagram showing prediction processing according to embodiment 1. FIG. 5 is a schematic diagram showing prediction result correction processing within the result determination correction processing according to embodiment 1. FIG. 6 is a schematic diagram showing parameter setting processing within the result determination correction processing according to embodiment 1. FIG. 7 is a flow diagram showing route determination processing according to embodiment 1. FIG. 8 is a schematic diagram showing route determination processing according to embodiment 1. FIG. 9 is a diagram showing an example of the functional configuration of a people flow measurement device according to a modification of embodiment 1. FIG. 10 is a diagram showing another example of the functional configuration of a people flow measurement device according to a modification of embodiment 1. FIG. 11 is a diagram showing another example of the hardware configuration of a people flow measurement device according to a modification of embodiment 1. FIG. 12 is a diagram showing an example of the functional configuration of a people flow measurement system according to embodiment 2. FIG. 13 is a flow diagram showing route model generation processing according to embodiment 2. FIG. 14 is a schematic diagram showing route model generation processing according to embodiment 3. FIG. 15 is a schematic diagram showing route model generation processing according to embodiment 3.

[0010] The present embodiment will be described below with reference to the drawings. In each drawing, identical or corresponding parts are designated by the same reference numerals. In the description of the embodiment, the description of identical or corresponding parts will be omitted or simplified as appropriate. Arrows in the drawings mainly indicate the flow of data or the flow of processing. Furthermore, the sized relationships between components in the following drawings may differ from the actual relationships. Furthermore, in the description of the embodiment, directions or positions such as up, down, left, right, front, rear, front, and back may be indicated. These notations are provided for the convenience of explanation and do not limit the placement, direction, or orientation of devices, instruments, or parts.

[0011] Embodiment 1. *** Description of Configuration *** FIG. 1 is a diagram showing an example of the hardware configuration of a people flow measurement system 500 according to this embodiment. The people flow measurement system 500 includes a people flow measurement device 100 and a mobile object 200. The mobile object 200 includes a camera 210. The camera 210 captures images of the surrounding area and outputs video data 20. The mobile object 200 moves along a movement path in a measurement area having multiple individual areas. Furthermore, while moving along the movement path, the mobile object 200 captures video data 20 of the measurement area with the camera 210. Specifically, the mobile object 200 is a moving object such as an AMR, a PMV, or a drone. AMR is an abbreviation for Autonomous Mobile Robot. PMV is an abbreviation for Personal Mobility Vehicle. Alternatively, the camera 210 may be attached to a person such as a security guard traveling along the travel route.

[0012] The people flow measuring device 100 is a computer. The people flow measuring device 100 includes a processor 910 as well as other hardware such as a memory 921, an auxiliary storage device 922, an input / output interface 930, and a communication device 950. The processor 910 is connected to the other hardware via signal lines and controls the other hardware.

[0013] The people flow measurement device 100 includes, as functional elements, a measurement unit 110, a prediction unit 120, a result determination unit 130, a display correction unit 140, a parameter setting unit 150, an analysis unit 160, a route determination unit 170, and a storage unit 180. The measurement unit 110 includes an image acquisition unit 111 and a people counting unit 112. The functions of the measurement unit 110, the prediction unit 120, the result determination unit 130, the display correction unit 140, the parameter setting unit 150, the analysis unit 160, and the route determination unit 170 are realized by software. The storage unit 180 stores a result database 81 and a congestion prediction model 82. The congestion prediction model 82 includes a route model 821 and a structure model 822. The storage unit 180 is included in a memory 921. The storage unit 180 may be included in an auxiliary storage device 922, or may be distributed between the memory 921 and the auxiliary storage device 922.

[0014] In the following description, the functions of the measurement unit 110, prediction unit 120, result determination unit 130, display correction unit 140, parameter setting unit 150, analysis unit 160, and route determination unit 170 may be referred to as functions of the people flow measurement device 100. Furthermore, the measurement unit 110, prediction unit 120, result determination unit 130, display correction unit 140, parameter setting unit 150, analysis unit 160, and route determination unit 170 may be referred to as each unit of the people flow measurement device 100.

[0015] The processor 910 is a device that executes a people flow measurement program. The people flow measurement program is a program that realizes the functions of the people flow measurement device 100. The processor 910 is an IC that performs arithmetic processing. Specific examples of the processor 910 are a CPU, a DSP, and a GPU. IC is an abbreviation for Integrated Circuit. CPU is an abbreviation for Central Processing Unit. DSP is an abbreviation for Digital Signal Processor. GPU is an abbreviation for Graphics Processing Unit.

[0016] The memory 921 is a storage device that temporarily stores data. Specific examples of the memory 921 are SRAM and DRAM. SRAM is an abbreviation for Static Random Access Memory. DRAM is an abbreviation for Dynamic Random Access Memory. The auxiliary storage device 922 is a storage device that saves data. A specific example of the auxiliary storage device 922 is an HDD. The auxiliary storage device 922 may also be a portable storage medium such as an SD (registered trademark) memory card, CF, NAND flash, flexible disk, optical disk, compact disk, Blu-ray (registered trademark) disk, or DVD. Note that HDD is an abbreviation for Hard Disk Drive. SD (registered trademark) is an abbreviation for Secure Digital. CF is an abbreviation for CompactFlash (registered trademark). DVD is an abbreviation for Digital Versatile Disk.

[0017] The input / output interface 930 is an interface for connecting input / output devices. Specific examples of the input / output interface 930 are USB and HDMI (registered trademark) ports. USB is an abbreviation for Universal Serial Bus. HDMI (registered trademark) is an abbreviation for High-Definition Multimedia Interface. Specific examples of input devices are devices such as a mouse, keyboard, or touch panel. Specific examples of output devices are display devices such as a monitor, and devices such as a printer.

[0018] The communication device 950 has a receiver and a transmitter, and performs wired or wireless communication. The communication device 950 is connected to a communication network such as a LAN, the Internet, a telephone line, or Wi-Fi (registered trademark). Specifically, the communication device 950 is a communication chip or NIC. NIC is an abbreviation for Network Interface Card.

[0019] The people flow measurement program is executed in the people flow measurement device 100. The people flow measurement program is loaded into the processor 910 and executed by the processor 910. The memory 921 stores not only the people flow measurement program but also an OS. OS is an abbreviation for Operating System. The processor 910 executes the people flow measurement program while running the OS. The people flow measurement program and the OS may be stored in an auxiliary storage device 922. The people flow measurement program and the OS stored in the auxiliary storage device 922 are loaded into the memory 921 and executed by the processor 910. Note that part or all of the people flow measurement program may be incorporated into the OS.

[0020] The people flow measurement device 100 may include multiple processors that replace the processor 910. These multiple processors share the task of executing the people flow measurement program. Each processor is a device that executes the people flow measurement program, just like the processor 910.

[0021] Data, information, signal values ​​and variable values ​​used, processed or output by the people flow measurement program are stored in memory 921, secondary storage device 922, or registers or cache memory within processor 910.

[0022] The "part" of each part of the people flow measuring device 100 may be read as a "circuit," "process," "procedure," "process," or "circuitry." The people flow measuring program causes a computer to execute each process, where the "part" of each part of the people flow measuring device 100 is read as a "process." That is, the people flow measuring program causes a computer to execute a measurement process, a prediction process, a result determination process, a display correction process, a parameter setting process, an analysis process, and a route determination process. The "process" of each process of the people flow measuring device 100 may be read as a "program," "program product," "computer-readable storage medium storing a program," or "computer-readable recording medium recording a program." Furthermore, the people flow measuring method is a method performed by the people flow measuring device 100 executing the people flow measuring program. The people flow measuring program may be provided by being stored in a computer-readable recording medium. Furthermore, the people flow measuring program may be provided as a program product.

[0023] ***Function Description*** Figure 2 is a diagram showing an example of the functional configuration of a people flow measurement system 500 according to this embodiment. The people flow measurement system 500 includes a people flow measurement device 100 and a mobile object 200. The mobile object 200 includes a camera 210. The mobile object 200 moves along a movement path 31 in a measurement area having multiple individual areas. The mobile object 200 also captures video data 20 of the measurement area with the camera 210 while moving along the movement path. The people flow measurement device 100 acquires the video data 20 from the mobile object 200 via a communication device 950.

[0024] The people flow measurement system 500 has a prediction function using the measurement unit 110 and the prediction unit 120, and a route determination function using the analysis unit 160 and the route determination unit 170 to determine the route of the moving object 200. Furthermore, the people flow measurement system 500 has a result determination correction function using the result determination unit 130, the display correction unit 140, and the parameter setting unit 150. The result determination correction function is a function that displays congestion more appropriately and optimizes the parameters of the route model 821 so that the route model 821 reflects actual measured values.

[0025] <Prediction Function> The video acquisition unit 111 of the measurement unit 110 acquires video data 20. The people counting unit 112 of the measurement unit 110 uses the video data 20 to count the number of people in each of the multiple individual areas as actual measurement results 21.

[0026] The prediction unit 120 uses the actual measurement results 21 and the congestion prediction model 82 to predict the number of people in each of the multiple individual areas as a prediction result 22. The congestion prediction model 82 includes a route model 821 and a structure model 822. The route model 821 is a model of the path of people flow in the measurement area. The route model 821 is a model that represents the paths that people take in the measurement area and branch off. The structure model 822 represents the structure of the measurement area. The structure model 822 is a model that represents the structure of the measurement area, such as map information, structure information, facility information, and area of ​​the measurement area.

[0027] Specifically, the prediction unit 120 uses the actual measurement results 21 and the congestion prediction model 82 to calculate the prediction results 22 using a multi-agent simulator. The multi-agent simulator is also called MAS. The result database 81 stores the actual measurement results 21 and the prediction results 22.

[0028] <Route Determination Function> The analysis unit 160 identifies the individual area with the lowest prediction accuracy among the multiple individual areas as a specific area based on the prediction results 22 and the actual measurement results 21, and estimates a people flow correlated with congestion in the specific area as a specific people flow. Then, the route determination unit 170 determines a movement route 31 for the mobile object 200 to acquire video data of the specific people flow.

[0029] <Result Determination and Correction Function> The result determination unit 130 determines whether the difference between the prediction result 22 and the actual measurement result 21 is equal to or greater than a threshold. The threshold is set appropriately in advance and stored in the storage unit 180. The display correction unit 140 corrects the prediction result 22 to the actual measurement result 21 if the difference between the prediction result 22 and the actual measurement result 21 is equal to or greater than the threshold. Then, the display correction unit 140 displays the congestion status of the measurement area on a display device based on the prediction result 22. Specifically, the display correction unit 140 displays the congestion status of each of the multiple individual areas on the display device. The display device is also called a congestion degree display device that displays the congestion status of the measurement area.

[0030] Furthermore, if the difference between the prediction result 22 and the actual measurement result 21 is equal to or greater than a threshold, the parameter setting unit 150 sets or updates the parameters of the route model 821 so that the route model 821 reflects the content of the actual measurement result 21. Note that the threshold used to determine whether to correct the prediction result 22 to the actual measurement result 21 may be different from the threshold used to determine whether to reflect the content of the actual measurement result 21 in the route model 821. For example, the threshold used to determine whether to reflect the content of the actual measurement result 21 in the route model 821 may be smaller than the threshold used to determine whether to correct the prediction result 22 to the actual measurement result 21, or vice versa.

[0031] ***Explanation of Operation*** Next, the operation of the people flow measurement system 500 according to this embodiment will be described. The operating procedure of the people flow measurement system 500 corresponds to a people flow measurement method. Furthermore, the program that realizes the operation of the people flow measurement system 500 corresponds to a people flow measurement program that executes people flow measurement processing.

[0032] The people flow measurement process, which is the operation of the people flow measurement system 500, includes a prediction process by the measurement unit 110 and the prediction unit 120, and a route determination process by the analysis unit 160 and the route determination unit 170 to determine the route of the moving object 200. Furthermore, the people flow measurement process includes a result determination correction process by the result determination unit 130, the display correction unit 140, and the parameter setting unit 150. The result determination correction process is a process that displays congestion more appropriately and optimizes the parameters of the route model 821 so that the route model 821 reflects actual measured values.

[0033] FIG. 3 is a flow diagram showing the prediction process and the result determination correction process in the people flow measurement process according to this embodiment.

[0034] <Prediction Process> In the prediction process, the video acquisition unit 111 acquires video data 20 output from the moving bodies 200 via the communication device 950. Specifically, each of the multiple moving bodies 200 captures an image of the measurement area while moving along a movement path 31 in the measurement area, and transmits the video data 20 to the people flow measurement device 100. The video acquisition unit 111 receives the video data 20 from each of the multiple moving bodies 200.

[0035] In step S101, the people counting unit 112 counts the number of people appearing in the video data 20. Specifically, the people counting unit 112 measures the actual measurement results for Tn using the video data 20 transmitted from multiple moving objects 200, where n is a natural number. The actual measurement results for Tn are the number of people in each of multiple individual areas included in the measurement area for Tn.

[0036] In step S102, the prediction unit 120 predicts the congestion at Tn+1 using the actual measurement results at Tn as input values. Specifically, the prediction unit 120 predicts the number of people in each individual area at Tn+1 as a prediction result using the actual measurement results at Tn and the congestion prediction model 82. More specifically, the prediction unit 120 predicts the number of people in each individual area at Tn+1 as a prediction result using the actual measurement results 21 at Tn and the congestion prediction model 82 with a multi-agent simulator.

[0037] When time has passed from Tn to Tn+1 (step S103), the people counting unit 112 uses the video data 20 to count the number of people in each individual area at Tn+1 as an actual measurement result.

[0038] <Result Determination and Correction Process: Steps S105 to S111> In step S105, the result determination unit 130 compares the actual measurement result at Tn+1 with the predicted result at Tn+1 to determine whether there is a difference. Specifically, it determines whether the difference between the predicted result at Tn+1 and the actual measurement result at Tn+1 is equal to or greater than a threshold. If the difference between the predicted result at Tn+1 and the actual measurement result at Tn+1 is equal to or greater than the threshold, the process proceeds to step S106. If the difference between the predicted result at Tn+1 and the actual measurement result at Tn+1 is smaller than the threshold, the process proceeds to step S108.

[0039] If it is determined that there is a difference between the actual measurement result at Tn+1 and the predicted result at Tn+1, the display correction unit 140 replaces the predicted value with the measured value in step S106. Specifically, if the difference between the actual measurement result at Tn+1 and the predicted result at Tn+1 is equal to or greater than a threshold, the display correction unit 140 replaces the predicted result at Tn+1 with the actual measurement result at Tn+1 to correct the predicted result at Tn+1.

[0040] If it is determined that there is a difference between the actual measurement result at Tn+1 and the predicted result at Tn+1, in step S107, the parameter setting unit 150 corrects the parameters near the individual area determined to have a difference, and reflects the actual measurement result in the congestion prediction model 82. Specifically, if the difference between the actual measurement result at Tn+1 and the predicted result at Tn+1 is equal to or greater than a threshold, the parameter setting unit 150 determines the parameters of the route model 821 so that the route model 821 reflects the contents of the actual measurement result. The parameters of the route model 821 are, for example, elements such as the process to the destination and the interactions between agents.

[0041] As described above, the threshold value used to determine whether to correct the predicted result at Tn+1 to the actual measurement result at Tn+1 and the threshold value used to determine whether to reflect the contents of the actual measurement result at Tn+1 in the route model 821 may be different values.

[0042] In step S108, n=n+1 is set, and the process proceeds to step S109.

[0043] In step S109, the display correction unit 140 displays the congestion status of each of the multiple individual areas based on the prediction result for Tn+1. The display correction unit 140 may display the congestion status of each of the multiple individual areas on a display device connected to the people flow measurement device 100 via the output interface. Alternatively, the display correction unit 140 may transmit the prediction result to another device via the communication device 950. The other device is a device that visualizes the congestion level of the measurement area based on the prediction result.

[0044] In step S110, it is determined whether the congestion display time has ended. If the congestion display time has not ended, the process proceeds to step S111. If the congestion display time has ended, the process ends.

[0045] In step S111, the prediction unit 120 predicts congestion at Tn+1 using the actual measurement result at Tn as an input value, and obtains a prediction result for Tn+1. Then, the process returns to step S103 and repeats the process. The prediction method in step S111 is the same as the prediction method in step S102.

[0046] Fig. 4 is a schematic diagram showing the prediction process according to this embodiment. Fig. 5 is a schematic diagram showing the display correction process of the result determination correction process according to this embodiment. Fig. 6 is a schematic diagram showing the parameter setting process of the result determination correction process according to this embodiment. A specific example of people flow measurement process will be described using Figs. 4 to 6. Using Figs. 4 to 6, black circles represent actual measurement results of the number of people, and white circles represent predicted results of the number of people. Actual measurement results are also called actual measurement values. Prediction results are also called predicted values.

[0047] As shown in (1) of Figure 4, a mobile object 200 acquires video data 20, which is people flow data, while moving through a measurement area 300. In Figure 4, the measurement area 300 is composed of multiple individual areas. Specifically, the individual areas refer to areas A, B, C, D, E, and F, as well as the routes connecting these areas. Alternatively, the individual areas may refer to each of areas A, B, C, D, E, and F.

[0048] As shown in Figure 4 (2), a people flow simulation using MAS is performed at Tn to predict the people flow and congestion in the measurement area at Tn+1. In Figure 4 (2), the number of people in areas A, B, C, D, E, and F, as well as in the routes connecting these areas, are obtained as prediction results for Tn+1.

[0049] As shown in (3) of FIG. 5, the predicted result at Tn+1 is compared with the actual measurement result at Tn+1, and if they differ, they are corrected. The corrected results are then displayed. As shown in the left diagram of FIG. 5, the actual measurement result for area D at Tn+1 is 6 people. On the other hand, as shown in the right diagram of FIG. 4, the predicted result for area D at Tn+1 is 1 person. The threshold is assumed to be 4 people. In this case, the result determination unit 130 determines whether the difference between the predicted result at Tn+1 (1 person) and the actual measurement result at Tn+1 (6 people) is equal to or greater than the threshold of 4 people. Because the difference between the predicted result and the actual measurement result is equal to or greater than the threshold, the display correction unit 140 corrects the predicted result at Tn+1 (1 person) to the actual measurement result at Tn+1 (6 people).

[0050] Furthermore, because the difference between the predicted and actual measurement results is equal to or greater than the threshold, the parameter setting unit 150 modifies the parameters of the route model in the simulation to be performed at Tn+1 based on the actual measurement results at Tn+1. Specifically, as shown in FIG. 6, the modifications are made as follows. For example, because the value for area D differed greatly between the predicted value and the actual measurement value, the movement branching rate, which is one of the parameters of the MAS, is modified as follows: - Increase the movement branching rate for people between A and D. - Decrease the movement branching rate for people between D and E. The parameters are changed in this way and used in the next simulation.

[0051] Data assimilation can also be used as a method for automatically adjusting parameters. Data assimilation is a method for incorporating measured data into numerical simulations. It is a method for determining more plausible parameters by finding values ​​that minimize the difference between the simulation output and the observed data.

[0052] 7 is a flow diagram showing the route determination process according to this embodiment. The route determination process is a process for appropriately determining the travel route of the moving object 200. The route determination process is performed after the prediction process has been performed and the prediction results and actual measurement results have been sufficiently accumulated in the result database 81. Therefore, the route determination process is a process that is performed asynchronously with the result determination correction process.

[0053] <Route Determination Process> In step S201, the analysis unit 160 identifies, as a specific area, an individual area with the lowest prediction accuracy among multiple individual areas based on the prediction results and actual measurement results stored in the result database 81. In step S202, the analysis unit 160 estimates, as a specific people flow, a people flow that is correlated with congestion in the specific area. Specifically, the analysis unit 160 performs a correlation analysis to identify locations that are influencing the congestion in the identified specific area. In step S203, the analysis unit 160 estimates, as a specific people flow, a people flow measurement point that is most correlated with congestion in the specific area.

[0054] In step S204, the route determination unit 170 determines a travel route 31 suitable for the moving object 200 to acquire video data of the specific people flow. Specifically, the route determination unit 170 determines the travel route 31 so as to pass through the location of the specific people flow with priority. In step S205, the route determination unit 170 transmits the travel route 31 to each moving object.

[0055] FIG. 8 is a schematic diagram illustrating a route determination process according to this embodiment. Specifically, the optimization of the movement route 31 of the moving object 200 is performed in the following procedure. In (11) of FIG. 8, the analysis unit 160 identifies area E, where the difference between the actual measurement results and the predicted results is large, as a specific area. In (12) of FIG. 8, the analysis unit 160 estimates the people flow that will affect the congestion in area E from the actual measurement results and predicted results acquired so far using a technique such as correlation analysis. From the data acquired so far, a correlation analysis with the congestion in area E is performed to estimate which people flow has the greatest impact. In (12) of FIG. 8, the route determination unit 170 sets the movement route 31 so as to prioritize the acquisition of estimated people flow data.

[0056] ***Other Configurations*** <Variation 1> In this embodiment, the people flow measuring device 100 has a configuration that can execute prediction processing, result determination correction processing, and route determination processing. On the other hand, the people flow measuring device 100 may be configured to execute only prediction processing and result determination correction processing. Furthermore, the people flow measuring device 100 may be configured to execute only prediction processing and route determination processing.

[0057] 9 is a diagram showing an example of the functional configuration of a people flow measuring device 100 according to a variation of this embodiment. The people flow measuring device 100 shown in FIG. 9 performs only the prediction process and the result determination and correction process. In other words, the analysis unit 160, the path determination unit 170, and the result database 81 may be omitted.

[0058] Fig. 10 is a diagram showing another example of the functional configuration of a people flow measuring device 100 according to a variation of this embodiment. The people flow measuring device 100 shown in Fig. 10 performs only prediction processing and route determination processing. In other words, the result determining unit 130, the display correcting unit 140, and the parameter setting unit 150 may be omitted. In the people flow measuring device 100 shown in Fig. 10, the prediction result 22 output by the prediction unit 120 is output to the congestion degree display device.

[0059] <Variation 2> In this embodiment, the functions of each unit of the people flow measuring device 100 are realized by software. As a variation, the functions of each unit of the people flow measuring device 100 may be realized by hardware. Specifically, the people flow measuring device 100 includes an electronic circuit 909 instead of the processor 910.

[0060] 11 is a diagram showing another example of the hardware configuration of the people flow measuring device 100 according to a modified example of this embodiment. The electronic circuit 909 is a dedicated electronic circuit that realizes the functions of each unit of the people flow measuring device 100. Specifically, the electronic circuit 909 is a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, a logic IC, a GA, an ASIC, or an FPGA. GA is an abbreviation for Gate Array. ASIC is an abbreviation for Application Specific Integrated Circuit. FPGA is an abbreviation for Field-Programmable Gate Array.

[0061] The functions of each part of the people flow measurement device 100 may be realized by a single electronic circuit, or may be distributed across multiple electronic circuits.

[0062] As another variation, some of the functions of each unit of the people flow measuring device 100 may be realized by electronic circuits, and the remaining functions may be realized by software. Also, some or all of the functions of each unit of the people flow measuring device 100 may be realized by firmware.

[0063] Each of the processor and electronic circuitry is also called a processing circuitry. In other words, the functions of each part of the people flow measurement device 100 are realized by the processing circuitry.

[0064] ***Explanation of the Effects of the Present Embodiment*** In a people flow simulation using MAS, a route model is set that determines the routes people will take. Therefore, it is necessary to either know people's behavior in the facility where people flow is being measured empirically, or to install sensors in multiple locations and observe them in advance. The people flow measurement system according to this embodiment can appropriately reflect actual measurement results in the route model, thereby achieving the effect of creating a route model with high accuracy and low cost. Note that the people flow measurement system according to this embodiment performs measurements using moving objects. However, the people flow measurement system according to this embodiment has the effect of being able to display congestion throughout the entire measurement target area by interpolating areas where no sensors are present using simulation results.

[0065] Furthermore, in a typical congestion display system, it is assumed that cameras are installed in locations where they can capture all the points that people pass through when moving from one area to another without any blind spots. Therefore, the cost of installing additional cameras is required for locations where cameras are not installed. On the other hand, the people flow measurement system according to this embodiment measures using moving objects, which has the effect of reducing the number of sensors and installation costs. Furthermore, the people flow measurement system according to this embodiment can focus on measuring data that is a factor in locations where prediction accuracy is low, thereby improving the accuracy of the congestion level.

[0066] Embodiment 2 In this embodiment, differences from and additions to embodiment 1 will be mainly described. In this embodiment, components having the same functions as those in embodiment 1 will be given the same reference numerals, and descriptions thereof will be omitted. In this embodiment, an aspect of automatically generating a route model will be described.

[0067] ***Configuration*** Fig. 12 is a diagram showing an example of the functional configuration of a people flow measurement system 500 according to this embodiment. In addition to the configuration of Fig. 9 described in embodiment 1, the people flow measurement device 100 according to this embodiment includes a movement trajectory accumulation unit 190, a movement trajectory database 101, and a model generation unit 102. Here, new functional elements described in this embodiment have been added to the configuration of Fig. 9 described in embodiment 1. Note that new functional elements described in this embodiment may also be added to the configuration of Fig. 2 or Fig. 10 described in embodiment 1.

[0068] The movement trajectory accumulation unit 190 accumulates the trajectory of a person moving within a measurement area as movement trajectory information in the movement trajectory database 101 based on the video data 20. The movement trajectory accumulation unit 190 includes a person detection unit 191, a person identification unit 192, and a movement trajectory calculation unit 193. The model generation unit 102 generates a route model 821 based on the movement trajectory information accumulated in the movement trajectory database 101.

[0069] ***Description of Operation*** In this embodiment, a route model generation process will be described in which a movement trajectory storage unit 190, a movement trajectory database 101, and a model generation unit 102 generate a route model 821 based on video data 20 acquired from a moving object 200.

[0070] 13 is a flow diagram showing the route model generation process according to this embodiment. In the route model generation process, a mobile object 200 is placed at the entrance of a measurement area, acquires video data 20 of the entrance, and transmits it to the people flow measurement device 100.

[0071] In step S301, the person detection unit 191 detects a person who has entered the measurement area as a person to be tracked, based on the video data 20 acquired from the moving object 200. In step S302, the person identification unit 192 identifies the person to be tracked based on the video data 20 acquired from the moving object 200, and tracks the person to be tracked until the person to be tracked exits the measurement area. In step S303, the movement trajectory calculation unit 193 registers the movement trajectory of the person to be tracked in the movement trajectory database 101. In step S304, the movement trajectory calculation unit 193 moves the moving object 200 to the entrance.

[0072] In step S305, the movement trajectory calculation unit 193 determines whether sufficient data has been acquired to create the route model 821. If it is determined that the data is not yet sufficient to create the route model 821, the process returns to step S301 and repeats the processing. If it is determined that sufficient data has been acquired to create the route model 821, the process proceeds to step S306.

[0073] In step S306, the model generation unit 102 tally the trajectories stored in the trajectory database 101. In step S307, the model generation unit 102 generates a route model 821 based on the tallying result of the trajectories. In step S308, the model generation unit 102 reflects the route model 821 in the congestion prediction model 82.

[0074] FIG. 14 is a schematic diagram showing the route model generation process according to this embodiment. Specifically, a route model is generated in the following procedure. In (21) of FIG. 14, a person who enters a measurement area is tracked by the moving object 200 until the person exits, and a movement trajectory is acquired. When a person is detected at the entrance, the moving object 200 may track the person to be tracked until the person exits. Alternatively, the person identification unit 192 may analyze video data from multiple moving objects 200 to track the person to be tracked from entry to exit.

[0075] 14 (21) is repeated until sufficient trajectory information is acquired, and then the model generation unit 102 creates a route model from the trajectories, as shown in FIG. 14 (22). Then, the route model 821 is reflected in the congestion prediction model 82. After that, the MAS is executed as described in the first embodiment.

[0076] ***Explanation of the Effects of This Embodiment*** As described above, the people flow measurement device according to this embodiment can automatically create a route model. Furthermore, by incorporating information on people tracking into the route model, it is possible to perform a simulation that is closer to the real environment, thereby improving estimation accuracy.

[0077] Embodiment 3 In this embodiment, differences from and additions to embodiment 2 will be mainly described. In this embodiment, components having the same functions as those in embodiment 2 will be assigned the same reference numerals, and descriptions thereof will be omitted.

[0078] *** Description of Configuration *** The configuration of the people flow measurement device 100 according to this embodiment is the same as that of embodiment 2. The movement trajectory accumulation unit 190 accumulates the trajectories of people moving through the measurement area for each attribute as movement trajectory information in the movement trajectory database 101. The model generation unit 102 generates a route model that takes into account the attributes of people moving through the measurement area, based on the movement trajectory information accumulated in the movement trajectory database 101.

[0079] 15 is a flow diagram showing the route model generation process according to this embodiment. In the route model generation process, as in the second embodiment, the mobile object 200 is placed at the entrance of the measurement area and outputs video data 20 of the entrance.

[0080] In step S401, the person detection unit 191 detects a person who has entered the measurement area as a person to be tracked, based on the video data 20 acquired from the moving object 200. At this time, the person detection unit 191 also detects the attributes of the person to be tracked. For example, the person detection unit 191 detects attributes such as whether the person to be tracked is male or female. In step S402, the person identification unit 192 identifies the person to be tracked based on the video data 20 acquired from the moving object 200, and tracks the person to be tracked until the person to be tracked exits the measurement area. In step S403, the movement trajectory calculation unit 193 accumulates the movement trajectory of the person to be tracked together with the attributes of the person to be tracked in the movement trajectory database 101. In step S404, the movement trajectory calculation unit 193 moves the moving object 200 to the entrance.

[0081] In step S405, the movement trajectory calculation unit 193 determines whether sufficient data has been acquired to create the route model 821. If it is determined that the data is not yet sufficient to create the route model 821, the process returns to step S401 and repeats the processing. If it is determined that sufficient data has been acquired to create the route model 821, the process proceeds to step S406.

[0082] In step S406, the model generation unit 102 aggregates the trajectories stored in the trajectory database 101 for each attribute. In step S407, the model generation unit 102 generates a route model 821 that takes the attributes into account, based on the result of aggregating the trajectories for each attribute. In step S408, the model generation unit 102 reflects the route model 821 in the congestion prediction model 82.

[0083] FIG. 16 is a schematic diagram showing the route model generation process according to this embodiment. Specifically, a route model that takes attributes into account is generated using the following procedure. In (31) of FIG. 16, when a person entering the measurement area is detected, the person's attributes are also detected. For example, it is detected whether the person to be tracked is male or female. The person is then tracked by the moving object 200 until they exit, and a movement trajectory for each attribute is obtained. When a person is detected at the entrance, the moving object 200 may track the person to be tracked until they exit. Alternatively, the person identification unit 192 may analyze video data from multiple moving objects 200 to track the person to be tracked from entry to exit.

[0084] After repeating step (31) in FIG. 16 and acquiring sufficient movement trajectory information, the model generation unit 102 creates a route model based on the acquired attributes and movement trajectories, as shown in step (32) in FIG. 16. Then, the route model 821 is reflected in the congestion prediction model 82. After that, the MAS is executed as described in the first embodiment. Note that while the attribute of male or female is detected in FIG. 16, other attributes such as age, number of people in a group, and whether the person is a family may also be detected.

[0085] ***Explanation of Effects of the Present Embodiment*** As described above, the people flow measurement device according to the present embodiment can automatically create a route model according to attributes. Furthermore, using information obtained by tracking people whose attributes have been detected, the congestion level can be estimated using the route model according to the attributes, thereby further improving estimation accuracy.

[0086] In the above first to third embodiments, each part of the people flow measurement device has been described as an independent functional block. However, the configuration of the people flow measurement device does not have to be the same as that of the above-described embodiments. The functional blocks of the people flow measurement device may have any configuration as long as they can realize the functions described in the above-described embodiments. Furthermore, the people flow measurement device may not be a single device, but may be a system composed of multiple devices. Furthermore, multiple parts of the first to third embodiments may be combined and implemented. Alternatively, only one part of these embodiments may be implemented. In addition, these embodiments may be combined in any way, either as a whole or in part. In other words, in the first to third embodiments, the embodiments may be freely combined, or any component of each embodiment may be modified, or any component of each embodiment may be omitted.

[0087] The above-described embodiments are essentially preferred examples and are not intended to limit the scope of the present disclosure, the scope of application of the present disclosure, or the scope of use of the present disclosure. The above-described embodiments can be modified in various ways as needed. For example, the procedures described using flow charts or sequence diagrams may be modified as appropriate.

[0088] 20 Video data, 21 Actual measurement results, 22 Prediction results, 31 Movement route, 81 Result database, 82 Congestion prediction model, 821 Route model, 822 Structural model, 100 People flow measurement device, 101 Movement trajectory database, 102 Model generation unit, 110 Measurement unit, 111 Image acquisition unit, 112 People counting unit, 120 Prediction unit, 130 Result determination unit, 140 Display correction unit, 150 Parameter setting unit, 160 Analysis unit, 170 Route determination unit, 180 Memory unit, 190 Movement trajectory accumulation unit, 191 Person detection unit, 192 Person identification unit, 193 Movement trajectory calculation unit, 200 Moving object, 210 Camera, 300 Measurement area, 500 People flow measurement system, 909 Electronic circuit, 910 Processor, 921 Memory, 922 Auxiliary storage device, 930 Input / output interface, 950 communication device.

Claims

1. A people flow measurement system comprising: a moving body that moves along a movement path in a measurement area having a plurality of individual areas, the moving body being equipped with a camera and acquiring video data of the measurement area using the camera; a measurement unit that uses the video data to measure the number of people in each of the plurality of individual areas as an actual measurement result; a prediction unit that predicts the number of people in each of the plurality of individual areas as a prediction result using the actual measurement results and a congestion prediction model including a route model that is a model of the route of people flow in the measurement area and a structural model that represents the structure of the measurement area; an analysis unit that identifies the individual area with the lowest prediction accuracy among the plurality of individual areas as a specific area based on the prediction results and the actual measurement results, and estimates a people flow that is correlated with congestion in the specific area as a specific people flow; and a route determination unit that determines the movement path for the moving body to acquire video data of the specific people flow.

2. The people flow measurement system according to claim 1, further comprising: a result determination unit which determines whether a difference between the predicted result and the actual measurement result is equal to or greater than a threshold; and a display correction unit which corrects the predicted result to the actual measurement result if the difference is equal to or greater than the threshold, and displays the congestion status in each of the plurality of individual areas based on the predicted result.

3. The people flow measurement system according to claim 2, further comprising a parameter setting unit that determines parameters of the route model so that the route model reflects the contents of the actual measurement results if the difference is equal to or greater than the threshold value.

4. The people flow measurement system according to any one of claims 1 to 3, comprising: a movement trajectory storage unit that stores trajectories of people moving through the measurement area in a movement trajectory database as movement trajectory information based on the video data; and a model generation unit that generates the route model based on the movement trajectory information stored in the movement trajectory database.

5. The people flow measurement system of claim 4, wherein the movement trajectory accumulation unit accumulates trajectories for each attribute of people moving through the measurement area in the movement trajectory database as the movement trajectory information, and the model generation unit generates the route model taking into account the attributes of people moving through the measurement area based on the movement trajectory information accumulated in the movement trajectory database.

6. A people flow measurement system as described in any one of claims 1 to 5, wherein the prediction unit calculates the prediction results using a multi-agent simulator, using the actual measurement results and the congestion prediction model.

7. A people flow measuring device comprising: a moving body moving along a movement path in a measurement area having a plurality of individual areas, the moving body being equipped with a camera, the measurement unit acquiring image data of the measurement area from the moving body using the camera, and measuring the number of people in each of the plurality of individual areas using the image data as an actual measurement result; a prediction unit predicting the number of people in each of the plurality of individual areas as a prediction result using the actual measurement result and a congestion prediction model including a route model that is a model of the route of people flow in the measurement area and a structural model that represents the structure of the measurement area; an analysis unit identifying an individual area among the plurality of individual areas with the lowest prediction accuracy as a specific area based on the prediction result and the actual measurement result, and estimating a people flow correlated with congestion in the specific area as a specific people flow; and a route determination unit that determines the movement path for the moving body to acquire the image data of the specific people flow.

8. A people flow measurement method in which a computer acquires image data from a moving body that moves along a movement path in a measurement area having a plurality of individual areas, the moving body being equipped with a camera and acquiring image data of the measurement area using the camera, and measures the number of people in each of the plurality of individual areas using the image data as an actual measurement result; the computer predicts the number of people in each of the plurality of individual areas as a prediction result using the actual measurement result and a congestion prediction model that includes a route model that is a model of the route of people flow in the measurement area and a structural model that represents the structure of the measurement area; the computer identifies the individual area among the plurality of individual areas with the lowest prediction accuracy as a specific area based on the prediction result and the actual measurement result, and estimates a people flow that is correlated with congestion in the specific area as a specific people flow; and the computer determines the movement path for the moving body to acquire the image data of the specific people flow.

9. A people flow measurement program that causes a computer to execute the following steps: a measurement process for acquiring image data from a moving body that moves along a movement path in a measurement area having a plurality of individual areas, the moving body being equipped with a camera and acquiring image data of the measurement area using the camera, and measuring the number of people in each of the plurality of individual areas as an actual measurement result using the image data; a prediction process for predicting the number of people in each of the plurality of individual areas as a prediction result using the actual measurement results and a congestion prediction model that includes a route model that is a model of the route of people flow in the measurement area and a structural model that represents the structure of the measurement area; an analysis process for identifying an individual area with the lowest prediction accuracy among the plurality of individual areas as a specific area based on the prediction results and the actual measurement results, and estimating a people flow that is correlated with congestion in the specific area as a specific people flow; and a route determination process for determining the movement path for the moving body to acquire the image data of the specific people flow.

Images