WO2022259541A1

WO2022259541A1 - Monitoring system, video quality setting method, crowd behavior analysis server, and computer-readable non-transitory recording medium for storing crowd behavior analysis program

Info

Publication number: WO2022259541A1
Application number: PCT/JP2021/022370
Authority: WO
Inventors: 妍王
Original assignee: 日本電気株式会社
Priority date: 2021-06-11
Filing date: 2021-06-11
Publication date: 2022-12-15
Also published as: JPWO2022259541A1

Abstract

This monitoring system (1) includes: a plurality of cameras (11-13, 21) arranged in a distributed manner within a management area; a crowd behavior analysis server (41) that analyzes, on the basis of video images acquired from the plurality of cameras (11-13, 21), the flow of people in each of individual monitoring areas respectively imaged by the plurality of cameras (11-13, 21) and that calculates a predicted people count value for predicting a change in the flow of people in each of the individual monitoring areas; and a video quality adjustment device (42) that sets the video image quality for a camera (21) corresponding to a relevant individual monitoring area such that the greater the predicted people count value, the higher the quality.

Description

Computer-readable non-temporary recording medium storing monitoring system, video quality setting method, crowd behavior analysis server, and crowd behavior analysis program

The present invention relates to a monitoring system, a video quality setting method, a crowd behavior analysis server, and a crowd behavior analysis program, and more particularly to a monitoring system and video quality setting method for analyzing the flow of people in a management area set over a wide range using a plurality of cameras. , crowd behavior analysis server and crowd behavior analysis program.

In places where people gather or where there is a strong flow of people, security guards and guides are placed to control the flow of people in order to avoid confusion. However, it is not possible to optimize the flow line and placement of guides without understanding the congestion situation and changes in the flow of people. Assigning personnel to monitor such changes in congestion and the flow of people not only requires more personnel, but also poses a problem in that it is difficult to accurately grasp the situation. Therefore, Patent Literature 1 discloses a congestion estimation system in which a plurality of cameras are arranged in a management area in order to grasp the congestion situation.

In the surveillance system described in Patent Document 1, when an analysis server analyzes low-quality surveillance video transmitted from a surveillance camera to detect an abnormality, and receives an image quality change request to change the image quality of the surveillance video to high quality, In addition, a bandwidth change request is sent to the bandwidth management device to change the bandwidth that can be used for transmission of surveillance video to a wider bandwidth, and an image quality change request is sent to the surveillance camera to send high-quality surveillance video. to send. As a result, it is possible to flexibly change the image quality of the monitoring video transmitted by the monitoring camera as needed without using unnecessary bandwidth.

Japanese Patent No. 6595287

However, with the monitoring system described in Patent Document 1, changes in video quality cannot follow rapid changes in the flow of people and the number of people, and there is a problem that appropriate monitoring cannot be performed in a management area where there are large changes in the flow of people and the number of people.

One aspect of the surveillance system according to the present invention includes a plurality of cameras dispersedly arranged in a management area, and each individual surveillance area captured by each of the plurality of cameras based on the images acquired from the plurality of cameras. A crowd behavior analysis server that analyzes the flow of people and calculates the number of people in each of the individual monitoring areas and a predicted number of people that predicts changes in the flow of people; and a video quality adjustment device for setting the quality to high.

One aspect of the video quality setting method for a surveillance system according to the present invention is a video quality setting method for a surveillance system that measures the flow of people in a management area using a plurality of cameras distributed and arranged in the management area. Then, based on the images acquired from the plurality of cameras, the number of people and the flow of people in each individual monitoring area captured by each of the plurality of cameras are analyzed, and a predicted number of people predicting changes in the flow of people in each of the individual monitoring areas is obtained. The video quality of the camera corresponding to the individual monitoring area is set higher as the predicted number of people is larger.

One aspect of the crowd behavior analysis server according to the present invention includes a video acquisition unit that acquires video of an individual monitoring area captured by each of a plurality of cameras dispersedly arranged in a management area; A people flow analysis unit that analyzes the flow of people that indicates the entry and exit and movement direction of people in each individual monitoring area, a crowd analysis unit that analyzes the number of people in each individual monitoring area based on the acquired video, and the flow of people in each individual monitoring area. a number prediction unit that calculates a number prediction value that is a value obtained by predicting a change and is used for switching the image quality of the camera, and the number prediction unit calculates the number of people in each individual monitoring area. Then, the predicted number of people is calculated by adding the number of people entering from the adjacent individual monitoring area and subtracting the number of people leaving the adjacent individual monitoring area.

One aspect of a computer-readable non-temporary recording medium in which a crowd behavior analysis program according to the present invention is stored is video of an individual monitoring area captured by each of a plurality of cameras dispersedly arranged in a management area. A computer-readable non-temporary recording medium in which a crowd behavior analysis program is stored based on the image acquisition processing for acquiring the images from the plurality of cameras, and the people in each individual monitoring area based on the acquired images People flow analysis processing for analyzing the flow of people showing entry/exit and direction of movement, Crowd analysis processing for analyzing the number of people in each individual monitoring area based on the acquired video, and values obtained by predicting changes in the flow of people in each individual monitoring area. a number-of-people prediction process for calculating a number-of-people prediction value used for switching the video quality of the camera, wherein the number-of-people prediction process calculates the number of people in each individual monitoring area, and calculates the number of people in the adjacent individual monitoring area. The predicted number of people is calculated by adding the number of people entering from and subtracting the number of people leaving the adjacent individual monitoring area.

According to the surveillance system, image quality setting method, crowd behavior analysis server, and crowd behavior analysis program according to the present invention, it is possible to reduce the amount of communication between the camera and the crowd analysis server while ensuring optimum image quality for people flow monitoring. can be done.

1 is a schematic diagram of a monitoring system according to a first embodiment; FIG. 1 is a block diagram of a monitoring system according to a first embodiment; FIG. 4 is a flowchart for explaining the operation of the monitoring system according to the first embodiment; 6 is a flowchart for explaining initialization processing of the monitoring system according to the first embodiment; FIG. 3 is a diagram illustrating an outline of calculation of a predicted number of people in the monitoring system according to the first embodiment; FIG. 7 is a flowchart for explaining camera image quality adjustment processing of the surveillance system according to the first embodiment; It is a figure explaining the people flow analysis process of step S22 of the monitoring system concerning Embodiment 1. FIG. It is a figure explaining the number-of-persons prediction process of step S23 of the monitoring system concerning Embodiment 1. FIG. FIG. 10 is a diagram illustrating video quality adjustment processing in step S24 of the monitoring system according to the first embodiment; FIG. It is a figure explaining the number-of-persons prediction process of step S23 of the monitoring system concerning Embodiment 2. FIG. It is a figure explaining the video quality adjustment process of step S23 of the monitoring system concerning Embodiment 3. FIG.

For clarity of explanation, the following descriptions and drawings have been omitted and simplified as appropriate. In addition, each element described in the drawing as a functional block that performs various processes can be configured with a CPU (Central Processing Unit), memory, and other circuits in terms of hardware, and memory implemented by a program loaded in the Therefore, those skilled in the art will understand that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof, and are not limited to either one. In each drawing, the same elements are denoted by the same reference numerals, and redundant description is omitted as necessary.

Also, the above-described program can be stored and supplied to the computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (e.g., flexible discs, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical discs), CD-ROMs (Read Only Memory), CD-Rs, CD-R/W, semiconductor memory (eg mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, RAM (Random Access Memory)). The program may also be delivered to the computer by various types of transitory computer readable media. Examples of transitory computer-readable media include electrical signals, optical signals, and electromagnetic waves. Transitory computer-readable media can deliver the program to the computer via wired channels, such as wires and optical fibers, or wireless channels.

Embodiment 1
FIG. 1 shows a schematic diagram of the monitoring system according to the first embodiment. FIG. 1 shows an example in which the monitoring system 1 is arranged in a predetermined management area. As shown in FIG. 1, the surveillance system 1 according to the first embodiment has cameras 11 to 13, 21, a public network 31, a crowd behavior analysis server 41, and a video quality adjustment device .

Then, in the example shown in FIG. 1,

cameras

11, 12, and 13 are arranged at the entrance/exit of the management area. Also, in the example shown in FIG. 1, the camera 21 is arranged at the branch point of the passage in the management area. In the monitoring system 1, at least cameras are arranged at the entrance/exit of the management area and at the junction of passages in the management area. By arranging the cameras in this manner, the monitoring system 1 improves the accuracy of the prediction of the number of people. Also, the number of cameras to be arranged in the monitoring system 1 may be appropriately set according to the size of the management area. For example, if there is a space where a crowd can gather in the management area, it is preferable to arrange the camera in the area where the crowd gathers. Further, in a straight portion of the passage in the management area where the moving speed of people falls within a certain range, it is preferable to widen the camera arrangement interval according to the moving speed of people. In this way, by arranging the cameras in accordance with the movement of people, the number of people can be predicted while reducing the number of installed cameras, that is, while reducing the number of images and the amount of data to be transmitted to the crowd behavior analysis server 41. Accuracy can be improved.

In addition, in the monitoring system 1, a video is captured by a camera installed in the management area, and the captured data is transmitted to the crowd behavior analysis server 41 via the public network 31. Therefore, reducing the amount of communication data leads to stable communication and a reduction in operating costs. At this time, in the monitoring system 1 according to the first embodiment, the crowd behavior analysis server 41 obtains a predicted number of people for each individual monitoring area monitored by each camera, and obtains a predicted number of people for each future individual monitoring area. By increasing the video quality according to the number of people, it reduces the amount of communication and improves the accuracy of people flow analysis. Therefore, the monitoring system 1 including the crowd behavior analysis server 41 and the video quality adjustment device 42 will be described in detail below.

FIG. 2 shows a block diagram of the monitoring system according to the first embodiment. FIG. 2 shows in detail the crowd behavior analysis server 41 among the components explained in FIG. Cameras 11 to 13 are cameras installed at the entrance/exit of the management area. Also, the camera 21 is a camera provided at a branch point of the passage within the management area. The cameras 11 to 13 and 21 respectively shoot moving images of individual monitoring areas set within the management area. This moving image is, for example, H.I. It is data compressed based on a predetermined standard such as H.264. In the monitoring system 1 , the video quality adjustment device 42 adjusts the quality of the image captured by the camera 21 based on the predicted number of people generated by the crowd behavior analysis server 41 . Assume that the higher the resolution of the captured moving image, the higher the quality. Further, although the details will be described later, in the monitoring system 1, the imaging quality of the cameras 11 to 13 arranged at the entrance/exit of the management area is fixed at the highest level. This is because the flow of people and the number of people at the entrance/exit are affected by the environment outside the management area, and this is to cope with unexpected environmental changes.

The crowd behavior analysis server 41 analyzes the flow of people in each individual monitoring area captured by each of the plurality of cameras (for example, cameras 11 to 13, 21) based on the images acquired from the cameras 11 to 13, 21, and Calculate the predicted number of people by predicting changes in the flow of people. The crowd behavior analysis server 41 adds the number of people entering from an adjacent individual monitoring area to the number of people in each individual monitoring area, and A number prediction value is calculated by subtracting the number of people leaving. That is, the crowd behavior analysis server 41 predicts the number of people in each individual monitoring area in the near future (for example, the next video acquisition cycle) based on the number of people entering and exiting each individual monitoring area measured up to the present time. Then, the video quality adjustment device 42 sets the video quality of the camera corresponding to the individual monitoring area higher as the predicted number of people increases.

Here, the crowd behavior analysis server 41 can be realized as dedicated hardware that performs image acquisition processing, crowd flow analysis processing, and number prediction processing, which will be described below. It can also be realized by The example shown in FIG. 2 shows the crowd behavior analysis server 41 that executes the crowd behavior analysis program for performing these processes on a computer.

As shown in FIG. 2, the crowd behavior analysis server 41 has a calculation unit 51, a memory 52, and a communication interface 53. The calculation unit 51 implements a video acquisition unit 61, a people flow analysis unit 62, a crowd analysis unit 63, and a number prediction unit 64 by executing a crowd behavior analysis program. The memory 52 is, for example, a storage unit using a volatile memory and a nonvolatile memory, and a crowd behavior analysis program is stored in the nonvolatile memory portion. A volatile memory portion of the memory 52 stores moving image data and calculation data used in calculations performed by the calculation unit 51 . The communication interface 53 performs communication processing for the crowd behavior analysis server 41 to communicate with the cameras 11 to 13 and 21 and the image quality adjustment device 42 . In the example shown in FIG. 2, the video quality adjustment device 42 is a separate device from the crowd behavior analysis server 41, but the video quality adjustment device 42 may be included as a function of the crowd behavior analysis server 41.

The video acquisition unit 61 performs video acquisition processing. In this video acquisition process, video of the individual monitoring area captured by each of a plurality of cameras dispersedly arranged in the management area is acquired. The people flow analysis unit 62 performs people flow analysis processing.

In the people flow analysis process, the people flow for each individual monitoring area is analyzed based on the acquired video. More specifically, in the people flow analysis process, the number of people who crossed the set boundary line surrounding the individual monitoring area and the number of people who entered the area were counted. Analyze which direction the person moved based on the position of

The crowd analysis unit 63 performs crowd analysis processing. In crowd analysis processing, the number of people in each individual monitoring area is counted based on the acquired video. More specifically, in the crowd analysis process, the number of people in each individual monitoring area during the first predetermined period of the video is counted, and the count value is used as the number of people.

The number of people prediction unit 64 performs the number of people prediction process. In the number-of-persons prediction process, a number-of-people prediction value, which is a value obtained by predicting a change in the number of people for each individual monitoring area and is used for switching the image quality of the camera, is calculated. Specifically, in the population prediction process, the number of people entering from an adjacent individual monitoring area is added to the number of people in each individual monitoring area, and the number of people leaving the adjacent individual monitoring area is calculated. By subtracting, the predicted number of people is calculated. It should be noted that the method of predicting the number of people by the people flow analysis unit 62 is not based on mere addition and subtraction, and prediction processing using artificial intelligence, for example, can also be used.

Next, the operation of the monitoring system 1 according to the first embodiment will be explained. Therefore, FIG. 3 shows a flowchart for explaining the operation of the monitoring system according to the first embodiment. As shown in FIG. 3, in the monitoring system 1 according to the first embodiment, after performing initialization processing (step S1) for initializing camera settings, people flow monitoring processing including camera image quality adjustment processing (step S2) is performed. I do. Note that FIG. 3 shows only the camera image quality adjustment process in the crowd monitoring process. Further, as shown in FIG. 3, in the surveillance system 1 according to the first embodiment, the camera image quality adjustment process is repeatedly executed in each preset periodic cycle (for example, a cycle with a set periodic time of 1 minute).

Here, the initialization process will be explained in detail. FIG. 4 shows a flowchart for explaining initialization processing of the monitoring system 1 according to the first embodiment. As shown in FIG. 4, in the initialization process, first, the image quality adjustment device 42 sets the image quality of the cameras (for example, cameras 11 to 13) installed at the entrance of the management area to the highest quality (step S11). Note that the camera for which the image quality is set in step S11 maintains the setting thereafter.

Next, the image quality adjustment device 42 sets the image quality of the camera (for example, camera 21) installed outside the entrance/exit of the management area to the lowest quality (step S12). It is preferable to perform the initialization process, for example, when there is no human intrusion into the management area or when there is extremely little intrusion. This is because under such conditions, it is easy to lower the imaging quality of cameras installed outside the entrance/exit of the management area to the lowest level.

Next, camera image quality adjustment processing in the surveillance system 1 according to the first embodiment will be described in detail. One of the features of this camera image quality adjustment process is to calculate the predicted number of people. Therefore, FIG. 5 is a diagram for explaining an outline of calculation of the predicted number of people in the monitoring system 1 according to the first embodiment.

As shown in FIG. 5, in the surveillance system 1 according to the first embodiment, each of the cameras 11 to 13, 21 shoots individual surveillance areas to generate moving images. Also, in FIG. 5, a solid line indicates a person who was in the individual monitoring area at the start of imaging in a certain imaging cycle, and a broken line indicates a person who entered the individual monitoring area at the end of the imaging cycle. That is, the number of people indicated by the solid line is the number of people calculated by the crowd analysis process, and the people indicated by the broken line are the people and the direction of movement of the people analyzed by the people flow analysis process. Then, in the monitoring system 1 according to the first embodiment, the number of people who left the individual monitoring area during the imaging cycle and the direction of movement are analyzed based on the number of people who were in the individual management area at the beginning of the imaging cycle. Then, based on this analysis result, a predicted value of the number of people predicted to exist in each individual monitoring area in the near future (for example, the next imaging cycle) is calculated as the predicted number of people.

In the example shown in FIG. 5, the individual management area of the camera 21 is the target of the predicted number of people, the number of people in the current shooting cycle is 5, the number of people leaving the area is 4, and the number of people entering the area is 3, and the number of people is predicted. The value is 4 people.

Next, camera image quality adjustment processing in the surveillance system 1 according to the first embodiment will be described in detail. FIG. 6 shows a flowchart for explaining camera image quality adjustment processing of the surveillance system according to the first embodiment.

As shown in FIG. 6, in the camera image quality adjustment process, the image acquisition unit 61 first receives captured images from each camera (step S21). Subsequently, the crowd analysis unit 62 analyzes the video received and calculates the number of people and the flow of people in each individual monitoring area, and the crowd analysis unit 63 analyzes the number of people in each individual monitoring area based on the video acquired. and a crowd analysis process are performed (step S22). Here, FIG. 7 shows a diagram for explaining the people flow analysis processing in step S22 of the monitoring system according to the first embodiment. In the example shown in FIG. 7, in the crowd analysis process in step S22, the initial number of people in the individual monitoring area photographed by the camera 11 is calculated to be 30 people. In the people flow analysis process, it is calculated that 20 people have moved toward the individual monitoring area photographed by the

camera

21 and 5 people have moved outside the management area. Also, in the crowd analysis process, the initial number of people in the individual monitoring area photographed by the camera 12 is calculated to be 14 people. In the people flow analysis process, it is calculated that after that, six people headed toward the individual monitoring area photographed by the camera 21, and five people moved out of the management area. Also, in the crowd analysis process, the initial number of people in the individual monitoring area photographed by the camera 13 is calculated to be nine. In the people flow analysis process, it is calculated that after that, three people headed toward the individual monitoring area photographed by the camera 21, and five people moved out of the management area. Also, in the crowd analysis process, the initial number of people in the individual monitoring area photographed by the camera 21 is calculated to be 10 people. In the people flow analysis process, after that, three people head toward the individual monitoring area captured by the camera 11, three people move toward the individual monitoring area set up by the camera 12, and four people move toward the individual monitoring area captured by the camera 13. It is calculated that it has moved in the direction of In this way, in the people flow analysis process of step S22, the initial number of people in each individual monitoring area, the number of people who moved after that, and the direction of movement are calculated.

Subsequently, in the camera image quality adjustment process, based on the calculation result of step S22, a people prediction process is performed to calculate a predicted number of people in each individual monitoring area in the next shooting cycle (step S23). Here, FIG. 8 is a diagram for explaining the number-of-people prediction processing in step S23 of the monitoring system 1 according to the first embodiment. As shown in FIG. 8, in the number prediction process, the number of people entering each individual monitoring area is added and the number of people leaving is subtracted to calculate the number prediction value. At this time, the monitoring system 1 according to the first embodiment does not calculate the predicted number of persons for the individual monitoring areas photographed by the cameras 11 to 13 at the entrances and exits of the management area. There are two reasons for the individual monitoring area: the influence of external factors that cannot be comprehended is large, resulting in large calculation errors, and the imaging quality of the cameras 11 to 13 is fixed. On the other hand, with respect to the camera 21 for which the predicted value of the number of people is to be calculated, the initial number of people is 29 when the number of people entering and exiting is applied.

Subsequently, in the camera image quality adjustment process, the image quality adjustment for adjusting the image quality setting of the camera 21 for capturing the individual monitoring area set other than the entrance/exit of the management area in the next image capturing cycle based on the calculation result of step S23. Processing is performed (step S24). Here, FIG. 9 shows a diagram for explaining the video quality adjustment processing in step S24 of the monitoring system 1 according to the first embodiment. As shown in FIG. 9, in the video quality adjustment process, a quality setting table that associates the predicted number of people with the video quality setting is referenced, and the predicted number of people calculated in step S23 is set to which quality setting table. Analyze whether it applies to the column. Then, the video quality setting value corresponding to the matched column is applied to the camera 21 . In the example shown in FIG. 9, the image quality of the camera 21 is changed from the lowest 640×360 to 1920×1080.

As described above, in the monitoring system 1 according to the first embodiment, each individual monitoring area in the near future is calculated using the number of people and the flow of people in adjacent individual monitoring areas among the current individual monitoring areas acquired using a plurality of cameras. A population prediction value is calculated by predicting the number of people in an area. In the surveillance system 1 according to the first embodiment, by increasing the image quality of the camera as the predicted number of people increases, the image quality is improved in advance and high followability is realized with respect to changes in the flow of people. However, the amount of data communication caused by transmission of video data can be reduced.

More specifically, in the monitoring system 1 according to Embodiment 1, the accuracy of the number of people calculation process in the crowd analysis process can be improved by preemptively improving the video quality based on the predicted number of people. Also, in the people flow analysis process, by improving the video quality, it is possible to improve the accuracy of grasping the number of people who have moved and the direction of movement. On the other hand, in the monitoring system 1 according to the first embodiment, even when the number of people and the flow of people decrease, the image quality can be lowered in advance, so the amount of communication can be reduced.

Embodiment 2
In a second embodiment, another form of the number-of-persons prediction process according to the first embodiment will be described. Therefore, FIG. 10 shows a diagram for explaining the process of predicting the number of people in step S23 of the monitoring system according to the second embodiment.

As shown in FIG. 10, in the number-of-persons prediction process according to the second embodiment, in consideration of people entering from outside the management area, individual Calculate the predicted number of people for the monitored area. Here, the number of people coming in from the outside is calculated based on past statistical data such as the day of the week and the time period. In addition, a value predicted by artificial intelligence can also be used for the number of inflows from the outside.

In this way, by calculating the estimated number of people in the individual monitoring area set at the entrance/exit of the management area, it is possible to set the resolution of the camera installed at the entrance/exit of the management area to an appropriate low level, thereby reducing the amount of data communication. can.

Embodiment 3
In the third embodiment, another form of setting items for video quality will be described. Therefore, FIG. 11 shows a diagram for explaining the video quality adjustment processing in step S23 of the monitoring system according to the third embodiment.

As shown in FIG. 11, in the monitoring system 1, the quality setting table can be set so that not only the resolution of the captured image but also the frame rate increases as the number of people in the individual monitoring area increases.

If the number of people increases, the accuracy of people flow analysis can be improved by increasing the frame rate. Therefore, by lowering the frame rate according to the number of people and the flow of people, it is possible to balance the accuracy of analysis of the flow of people and the amount of data communication.

It should be noted that the present invention is not limited to the above embodiments, and can be modified as appropriate without departing from the scope of the invention.

(Appendix 1)
a plurality of cameras distributed in a management area;
Analyze the number of people and the flow of people in each individual monitoring area captured by each of the plurality of cameras based on the images acquired from the plurality of cameras, and calculate a predicted number of people predicting changes in the flow of people in each of the individual monitoring areas. a crowd behavior analysis server;
a video quality adjustment device that sets the video quality of the camera corresponding to the individual monitoring area higher as the predicted number of people increases;
surveillance system.
(Appendix 2)
The crowd behavior analysis server adds the number of people entering from the adjacent individual monitoring area to the number of people in each individual monitoring area, and calculates the number of people leaving the adjacent individual monitoring area. The monitoring system according to appendix 1, wherein the predicted number of people is calculated by subtraction.
(Appendix 3)
3. The monitoring system according to appendix 1 or 2, wherein the crowd behavior analysis server does not calculate the predicted number of people in an individual monitoring area located at an entrance of the management area among the plurality of individual monitoring areas.
(Appendix 4)
The crowd behavior analysis server adds/subtracts a statistically calculated predicted entry/exit value to/from the number of people in the individual monitoring area for the individual monitoring area located at the entrance/exit of the management area among the plurality of individual monitoring areas. 3. The monitoring system according to appendix 1 or 2, wherein the predicted number of people is calculated by
(Appendix 5)
5. The surveillance system according to any one of appendices 1 to 4, wherein the video quality adjustment device sets a predetermined fixed image quality to a camera positioned at an entrance/exit of the management area among the plurality of cameras.
(Appendix 6)
5. The monitoring system according to any one of supplementary notes 1 to 4, wherein the image quality adjustment device sets a camera positioned at an entrance of the management area among the plurality of cameras to the highest image quality that can be set.
(Appendix 7)
The video quality adjustment device has a video quality table describing video quality corresponding to the predicted number of people, and refers to the video quality table to improve the video quality set in the camera as the predicted number of people increases. 7. The monitoring system according to any one of appendices 1 to 6, wherein
(Appendix 8)
8. The monitoring system according to any one of appendices 1 to 7, wherein the video quality adjustment device adjusts at least one of video resolution and frame rate as the video quality.
(Appendix 9)
9. The monitoring system according to any one of appendices 1 to 8, wherein the individual monitoring area is set at least at a branch point and an entrance/exit in the management area.
(Appendix 10)
10. The monitoring system according to any one of appendices 1 to 9, wherein the individual monitoring area is set to include an area in which people tend to stay in the management area.
(Appendix 11)
Supplementary note 1 that the individual monitoring areas are set such that the distance between the adjacent individual monitoring areas increases as the moving speed of the person increases in the portion where the person moves within the management area within a certain speed range. 11. A surveillance system according to any one of claims 1-10.
(Appendix 12)
A video quality setting method for a surveillance system for measuring the flow of people in a management area using a plurality of cameras dispersedly arranged in the management area,
Analyze the flow of people in each individual monitoring area captured by each of the plurality of cameras based on the images acquired from the plurality of cameras, and calculate a predicted number of people that predicts changes in the number of people and the flow of people in each of the individual monitoring areas. ,
A video quality setting method for a monitoring system, wherein the video quality of the camera corresponding to the individual monitoring area is set higher as the predicted number of people increases.
(Appendix 13)
a video acquisition unit that acquires video of an individual monitoring area captured by each of a plurality of cameras dispersedly arranged in a management area;
a people flow analysis unit that analyzes the flow of people showing the entry and exit and movement direction of people in each individual monitoring area based on the acquired video;
a crowd analysis unit that analyzes the number of people in each individual monitoring area based on the acquired video;
a number prediction unit that calculates a number prediction value that is a value obtained by predicting a change in the flow of people in each of the individual monitoring areas and is used for switching the image quality of the camera,
The number prediction unit adds the number of people entering from the adjacent individual monitoring area to the number of people in each individual monitoring area and subtracts the number of people leaving the adjacent individual monitoring area. A crowd behavior analysis server that calculates the predicted number of people by
(Appendix 14)
A computer-readable non-temporary recording medium in which a crowd behavior analysis program is stored based on images of an individual monitoring area captured by each of a plurality of cameras dispersedly arranged in a management area,
an image acquisition process for acquiring the images from the plurality of cameras;
People flow analysis processing for analyzing the flow of people showing the entry and exit and movement direction of people in each individual monitoring area based on the acquired video;
Crowd analysis processing for analyzing the number of people in each individual monitoring area based on the acquired video;
a number prediction process for calculating a number prediction value that is a value obtained by predicting a change in the flow of people for each of the individual monitoring areas and is used for switching the image quality of the camera,
The number of people prediction process adds the number of people entering from the adjacent individual monitoring area to the number of people in each individual monitoring area and subtracts the number of people leaving the adjacent individual monitoring area. A computer-readable non-temporary recording medium in which a crowd behavior analysis program for calculating the predicted number of people is stored.

1 Surveillance System 11 Camera 12 Camera 13 Camera 21 Camera 31 Public Network 41 Crowd Behavior Analysis Server 42 Image Quality Adjustment Device 51 Calculation Unit 52 Memory 53 Communication Interface 61 Image Acquisition Unit 62 People Flow Analysis Unit 63 Crowd Analysis Unit 64 People Prediction Unit

Claims

a plurality of cameras distributed in a management area;
Analyze the number of people and the flow of people in each individual monitoring area captured by each of the plurality of cameras based on the images acquired from the plurality of cameras, and calculate a predicted number of people predicting changes in the flow of people in each of the individual monitoring areas. a crowd behavior analysis server;
a video quality adjustment device that sets the video quality of the camera corresponding to the individual monitoring area higher as the predicted number of people increases;
surveillance system.
The crowd behavior analysis server adds the number of people entering from the adjacent individual monitoring area to the number of people in each individual monitoring area, and calculates the number of people leaving the adjacent individual monitoring area. 2. The monitoring system according to claim 1, wherein the predicted number of people is calculated by subtraction.
　The monitoring system according to claim 1 or 2, wherein the crowd behavior analysis server does not calculate the predicted number of people in an individual monitoring area located at the entrance/exit of the management area among the plurality of individual monitoring areas.
The crowd behavior analysis server adds/subtracts a statistically calculated predicted entry/exit value to/from the number of people in the individual monitoring area for the individual monitoring area located at the entrance/exit of the management area among the plurality of individual monitoring areas. 3. The monitoring system according to claim 1 or 2, wherein the predicted number of people is calculated by .
The monitoring system according to any one of claims 1 to 4, wherein the image quality adjustment device sets a predetermined fixed image quality to a camera positioned at the entrance/exit of the management area among the plurality of cameras.
The monitoring system according to any one of claims 1 to 4, wherein the image quality adjustment device sets a camera located at the entrance/exit of the management area among the plurality of cameras to the highest image quality that can be set.
The video quality adjustment device has a video quality table describing video quality corresponding to the predicted number of people, and refers to the video quality table to improve the video quality set in the camera as the predicted number of people increases. 7. The surveillance system according to any one of claims 1 to 6.
The monitoring system according to any one of claims 1 to 7, wherein the video quality adjustment device adjusts at least one of video resolution and frame rate as the video quality.
The monitoring system according to any one of claims 1 to 8, wherein the individual monitoring areas are set at least at branch points and entrances/exits in the management area.
The monitoring system according to any one of claims 1 to 9, wherein the individual monitoring areas are set to include areas in which people tend to stay in the management area.
3. The individual monitoring areas are set such that the distance between adjacent individual monitoring areas increases as the moving speed of the person increases in a portion where a person moves within the management area within a certain speed range. 11. The surveillance system according to any one of 1 to 10.
A video quality setting method for a surveillance system for measuring the flow of people in a management area using a plurality of cameras dispersedly arranged in the management area,
Analyze the flow of people in each individual monitoring area captured by each of the plurality of cameras based on the images acquired from the plurality of cameras, and calculate a predicted number of people that predicts changes in the number of people and the flow of people in each of the individual monitoring areas. ,
A video quality setting method for a monitoring system, wherein the video quality of the camera corresponding to the individual monitoring area is set higher as the predicted number of people increases.
a video acquisition unit that acquires video of an individual monitoring area captured by each of a plurality of cameras dispersedly arranged in a management area;
a people flow analysis unit that analyzes the flow of people showing the entry and exit and movement direction of people in each individual monitoring area based on the acquired video;
a crowd analysis unit that analyzes the number of people in each individual monitoring area based on the acquired video;
a number prediction unit that calculates a number prediction value that is a value obtained by predicting a change in the flow of people in each of the individual monitoring areas and is used for switching the image quality of the camera,
The number prediction unit adds the number of people entering from the adjacent individual monitoring area to the number of people in each individual monitoring area and subtracts the number of people leaving the adjacent individual monitoring area. A crowd behavior analysis server that calculates the predicted number of people by
A computer-readable non-temporary recording medium in which a crowd behavior analysis program is stored based on images of an individual monitoring area captured by each of a plurality of cameras dispersedly arranged in a management area,
an image acquisition process for acquiring the images from the plurality of cameras;
People flow analysis processing for analyzing the flow of people showing the entry and exit and movement direction of people in each individual monitoring area based on the acquired video;
Crowd analysis processing for analyzing the number of people in each individual monitoring area based on the acquired video;
a number prediction process for calculating a number prediction value that is a value obtained by predicting a change in the flow of people for each of the individual monitoring areas and is used for switching the image quality of the camera,
The number of people prediction process adds the number of people entering from the adjacent individual monitoring area to the number of people in each individual monitoring area and subtracts the number of people leaving the adjacent individual monitoring area. A computer-readable non-temporary recording medium in which a crowd behavior analysis program for calculating the predicted number of people is stored.