Monitoring device, monitoring method, and computer program

Abstract

To improve the performance of detecting abnormality in a monitoring area including a human flow.SOLUTION: An acquisition unit in a monitoring device acquires a moving image obtained by imaging a monitoring area as a monitoring moving image. An estimation unit uses information of the orientation of a crowd detected from the monitoring moving image to estimate an abnormality occurrence area in which abnormality has occurred in the monitoring area. An output unit outputs information indicating the abnormality occurrence area in the monitoring area.SELECTED DRAWING: Figure 10

Description

[Technical field]

[0001] The present disclosure relates to video surveillance technology. [Background technology]

[0002] Various techniques have been proposed for monitoring an area to be monitored using images captured of the area to be monitored. Patent Document 1 (International Publication No. 2002 / 021441) discloses a technique for estimating the location of an abnormality that caused the abnormal behavior of a moving object detected from an image captured in a monitored area when the moving object is determined to be behaving abnormally. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2002 / 021441 Summary of the Invention [Problem to be solved by the invention]

[0004] In Patent Document 1, when a moving object behaves abnormally within a monitored area, the abnormal behavior can be detected. However, there are various types of abnormalities that occur within a monitored area. For example, the behavior of each person within the monitored area is not abnormal, but there is also an abnormality in which a package suspected to be a dangerous item is placed within the monitored area. In this case, since the behavior of the person (moving object) is not abnormal, the technology shown in Patent Document 1 has difficulty detecting such an abnormality.

[0005] A main objective of the present disclosure is to provide a technology that can contribute to improving the performance of detecting abnormalities in a monitoring area with a flow of people (traffic), with regard to video surveillance. [Means for solving the problem]

[0006] In order to achieve the above object, a monitoring device according to the present disclosure includes, as one aspect thereof, an acquisition unit that acquires a video captured in a surveillance area as a surveillance video; an estimation unit that estimates an abnormality occurrence region where an abnormality has occurred in a surveillance area by using information on the direction of the crowd detected from the surveillance video; an output unit that outputs information indicating an abnormality occurrence area in a monitoring area; Equipped with.

[0007] In addition, as one aspect, a monitoring method according to the present disclosure includes: By computer, The video captured in the surveillance area is acquired as surveillance video. Using crowd direction information detected from surveillance video, anomaly occurrence areas in the surveillance area are estimated, Information indicating the area in the monitoring area where an abnormality has occurred is output.

[0008] Furthermore, in one aspect, the computer program according to the present disclosure comprises: A process of acquiring a video of a monitoring area as a monitoring video; A process of estimating an abnormality occurrence area in a monitoring area using crowd direction information detected from a monitoring video; A process of outputting information indicating an abnormality occurring area in a monitoring area; to be executed by the computer. Effect of the Invention

[0009] The present disclosure can contribute to improving the performance of video surveillance in detecting abnormalities in surveillance areas where there is a flow of people (traffic). [Brief description of the drawings]

[0010] [Figure 1] FIG. 2 is a diagram illustrating a configuration example of a monitoring device according to the present disclosure. [Diagram 2] FIG. 2 is a diagram illustrating an example of a monitoring area. [Diagram 3] 1 is a diagram showing a specific example of a detected crowd, the direction of the crowd, and an abnormality occurrence area. [Figure 4] 13 is a diagram showing another specific example of a detected crowd, the direction of the crowd, and an abnormality occurrence area. FIG. [Diagram 5] FIG. 11 is a diagram showing another specific example of a detected crowd, the direction of the crowd, and an abnormality occurrence area. [Figure 6] 11A and 11B are diagrams illustrating an example of a display mode for displaying abnormality occurrence region information. [Figure 7] 13A and 13B are diagrams illustrating another example of a display mode for displaying abnormality occurrence region information. [Figure 8] 13 is a diagram illustrating yet another example of a display mode for displaying abnormality occurrence region information. FIG. [Figure 9] 10 is a flowchart illustrating an example of the operation of a monitoring device according to the present disclosure. [Figure 10] FIG. 13 is a diagram illustrating another example configuration of a monitoring device according to the present disclosure. [Figure 11] 10 is a flowchart illustrating another example of the operation of the monitoring device according to the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0012] First Embodiment The monitoring device of the first embodiment is a device related to video monitoring that monitors a predetermined monitoring area using video captured in the monitoring area. This monitoring device is a computer device, has a configuration as shown in Fig. 1, and is connected to a photographing device 5 and a display device 7 by wire or wirelessly.

[0013] The photographing device 5 is, for example, a network camera, and is capable of photographing video. Here, the photographing device 5 is installed in an appropriate position where it can photograph the monitoring area as shown in FIG. 2. Note that, although one photographing device 5 is shown in the examples of FIG. 1 and FIG. 2, the number of photographing devices 5 photographing the monitoring area is not limited, and there may be more than one. When multiple photographing devices 5 are used, for example, the photographing devices 5 are installed in appropriate positions so as to photograph different areas in the monitoring area, or so that part of the photographing range of one photographing device 5 overlaps with the photographing range of another photographing device 5.

[0014] The display device 7 is a device that displays images, characters, and the like, and the display operation thereof is controlled by the monitoring device 1 here.

[0015] The monitoring device 1 includes a calculation device 10 and a storage device 20. The storage device 20 includes a storage medium for storing data and a computer program (hereinafter also referred to as a program) 21. There are multiple types of storage devices, such as a magnetic disk device and a semiconductor memory element, and there are multiple types of semiconductor memory elements, such as a RAM (Random Access Memory) and a ROM (Read Only Memory). A computer device includes multiple types of storage devices according to their applications, but these storage devices are collectively referred to as a storage device 20 without distinction here. Furthermore, the types and number of storage devices 20 included in the monitoring device 1 are not limited, and a description thereof will be omitted. Furthermore, the monitoring device 1 may write data to an external storage device (database) and read data from the storage device, but a description thereof will be omitted here even if data is read and written to an external storage device.

[0016] The arithmetic device 10 is configured with a processor such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit). The arithmetic device 10 can have various functions based on a program 21 by reading and executing the program 21 stored in a storage device 20. Here, the arithmetic device 10 has an acquisition unit 11, a detection unit 12, an estimation unit 13, and an output unit 14 as functional units related to video surveillance.

[0017] The acquisition unit 11 acquires a video of the monitoring area captured by the image capture device 5 as a monitoring video. Here, the monitoring video is output from the image capture device 5 to the monitoring device 1 immediately after being captured. As a result, the acquisition unit 11 acquires the monitoring video from the image capture device 5 in real time. The acquisition unit 11 also stores the acquired monitoring video in the storage device 20. The monitoring video output from the image capture device 5 is associated with information related to the capture, such as image capture time information related to the time of capture, and identification information for identifying the image capture device 5 that captured the video. When the acquisition unit 11 stores the monitoring video in the storage device 20, the video is stored in the storage device 20 in a state in which such information is associated with the video. When the monitoring video is stored in the storage device 20, the monitoring video is stored in the storage device 20 according to a predetermined storage method, such as storing the video for each predetermined time length unit.

[0018] In addition, various situations are assumed in which the acquisition unit 11 acquires surveillance videos, such as continuously, only during a set time period every day, or only at a preset date and time. For this reason, the surveillance device 1 may be provided with a configuration for receiving setting information related to the time period, date and time, etc., for acquiring such surveillance videos, and the acquisition unit 11 may execute a process for acquiring the surveillance videos using the setting information. There are various configurations for settings related to the acquisition of such surveillance videos, and here, for example, an appropriate configuration considering the type of the imaging device 5, etc. may be adopted, and a description thereof will be omitted.

[0019] The detection unit 12 detects a crowd in a monitoring area from a surveillance video and detects an abnormality in the monitoring area by determining whether the detected crowd is abnormal. For example, when the acquisition unit 11 acquires a surveillance video, the detection unit 12 sequentially processes the acquired surveillance video. For example, a crowd behavior analysis technique is used for such abnormality detection by the detection unit 12. In the crowd behavior analysis technique, for example, in each of the frame images to be analyzed among the multiple frame images constituting the surveillance video, the frame images are divided into meshes, and the feature amounts of the density and flow of people are extracted for each divided area. The frame images to be analyzed here are frame images analyzed by the crowd behavior analysis technique. The frame images to be analyzed may be all the frame images constituting the surveillance video, or may be frame images for a predetermined number of frames, such as every 10 frames, among the frame images in the time series constituting the surveillance video. The frame images to be analyzed are appropriately set in advance from the multiple frame images constituting the surveillance video.

[0020] Then, the crowd and the direction of the crowd in the surveillance video are detected from the temporal changes in the extracted feature amounts (in other words, the flow of people). It is also determined whether the state of the crowd is abnormal or not. The direction of the crowd here refers to the main direction in which the people making up the crowd are facing. Crowd behavior analysis technology does not detect the direction of each person making up the crowd, but rather detects the direction in which the crowd as a whole is facing.

[0021] Furthermore, when a detected crowd is in an abnormal state, the crowd behavior analysis technology can determine the type of abnormal state, such as abnormal congestion, a crowd dispersing due to escape, a surrounded state, or abnormal accumulation.

[0022] Incidentally, a predetermined criterion for judging the occurrence of an abnormality is used in the process of judging whether or not the state of the crowd in the monitoring area is abnormal by the detection unit 12. The criterion for judging the occurrence of an abnormality is set, for example, based on a standard state of people flow in the monitoring area obtained by monitoring the monitoring area. Here, for example, assume that the number of people flowing in the monitoring area in the standard state is about 500 people. In such a case, if the process of judging the occurrence of an abnormality in the monitoring area is performed using the criterion for judging the occurrence of an abnormality set based on the number of people flowing in the standard state, there may be a concern that the reliability of the judgment result will decrease.

[0023] In the first embodiment, in consideration of such concerns, the detection unit 12 further includes the following function. That is, as a criterion for judging the occurrence of an abnormality in the monitoring area, in addition to a criterion for judging the occurrence of an abnormality determined based on a pedestrian flow in a standard state (hereinafter also referred to as a standard criterion), at least one criterion for judging the occurrence of an abnormality determined based on a pedestrian flow less than the standard state (hereinafter also referred to as a criterion for judging a low pedestrian flow state) is set. The criterion for judging a low pedestrian flow state is set so that the occurrence of an abnormality is more difficult to detect than when the standard criterion is used, in order to suppress a decrease in the reliability of the judgment result. In addition, information representing the number of people flowing in the standard state in the monitoring area is stored, for example, in the storage device 20. Here, the information representing the number of people flowing in the standard state is information representing a range of the number of people flowing in the standard state, for example, 450 to 550 people.

[0024] The detection unit 12 detects the number of people flowing in the monitoring area from the surveillance video as the detected number of people, and compares the detected number of people with a threshold. The threshold here is a numerical value set using information representing the number of people flowing in a standard state, and for example, a lower limit value of the range of the number of people flowing in the standard state is set. The process of comparing the detected number of people with such a threshold can also be said to be a process of comparing the flow of people in a predetermined standard state in the monitoring area with the flow of people detected from the surveillance video.

[0025] When the number of detected people is equal to or greater than the threshold, the detection unit 12 determines to use the standard judgment criteria, and executes the judgment process of the occurrence of an abnormality in the monitoring area using the standard judgment criteria. When the number of detected people is less than the threshold, the detection unit 12 determines to use the judgment criteria of the low-people flow state, and executes the judgment process of the occurrence of an abnormality in the monitoring area using the judgment criteria of the low-people flow state. That is, the detection unit 12 detects the occurrence of an abnormality in the monitoring area using the result of comparing the people flow in the standard state with the people flow detected from the monitoring video. Note that the range of the number of people flowing from zero to the threshold may be divided into multiple ranges, and the judgment criteria of the low-people flow state may be set for each divided range of the number of people (hereinafter also referred to as the number of people division range). In this case, when the number of detected people is less than the threshold, the detection unit 12 determines to use the judgment criteria of the low-people flow state set for the number of people division range that includes the detected number of people.

[0026] When the detection unit 12 detects that an abnormality has occurred in the monitoring area, the estimation unit 13 estimates an abnormality occurrence region in which an abnormality has occurred in the monitoring area, using information on the direction of the crowd detected from the monitoring video. For example, as shown in FIG. 3, it is assumed that a plurality of crowds G1 to G5 are detected in the monitoring area, and each of the crowds G1 to G5 is in an abnormally staying state, and an abnormality is detected in the monitoring area. It is further assumed that the crowds G1 to G5 are in an abnormally surrounding state. It is further assumed that the direction of the crowds as represented by the arrows A1 to A5 in FIG. 3 is detected for each of the crowds G1 to G5. The estimation unit 13 estimates an abnormality occurrence region S, using information on the directions (directions of the arrows) A1 to A5 of the crowds G1 to G5 in the abnormally staying state. That is, it is assumed that an abnormally staying crowd has occurred, and some abnormality is occurring in the direction in which the crowd is facing. From this, the estimation unit 13 estimates the abnormality occurrence region S from the directions A1 to A5 in which the abnormally staying crowds G1 to G5 are facing.

[0027] In the example of FIG. 3, a case where a plurality of abnormally staying crowds G1 to G5 are detected is shown, but for example, a single abnormally staying crowd G in a surrounding state as shown in FIG. 4 may be detected. In such a case, for example, when a plurality of directions A are detected as the direction of the crowd as shown in FIG. 4, the estimation unit 13 estimates an abnormal occurrence area S as shown in FIG. 4 from the direction A of the crowd. Furthermore, as shown in FIG. 5, a case may be detected where a plurality of abnormally staying crowds G10 to G12 are detected, and the crowds G10 to G12 are not in a surrounding state. In such a case, for example, when directions A10 to A12 as shown in FIG. 5 are detected as the direction of each of the crowds G10 to G12, the estimation unit 13 estimates an abnormal occurrence area S as shown in FIG. 5 from the direction A10 to A12 of the crowd.

[0028] Furthermore, in the above example, when an abnormality in the monitoring area is detected by detecting an abnormally staying crowd, the estimation unit 13 estimates the abnormality occurrence area S by using information on the direction of the crowd. The estimation unit 13 may also estimate the abnormality occurrence area S when, for example, a crowd in an abnormal state is detected in which the crowd is dispersing due to escape. In such a case, the direction of the crowd is set to the direction in which the crowd is dispersing. The estimation unit 13 estimates the original area of ​​such crowd direction (direction in which the crowd is dispersing) as the abnormality occurrence area S.

[0029] The output unit 14 outputs the monitoring video (real-time video) captured by the imaging device 5 to the display device 7. When the estimation unit 13 estimates an abnormality occurrence area in the monitoring area, the output unit 14 outputs information representing the abnormality occurrence area as abnormality occurrence area information to the display device 7. Here, the output unit 14 also has a function of controlling the display of the display device 7. That is, the monitoring video and the abnormality occurrence area information are displayed on the display screen of the display device 7 by the display control of the output unit 14. The display mode of the monitoring video and the abnormality occurrence area information on the display screen of the display device 7 is determined in advance. For example, one example of the display mode is shown in FIG. 6.

[0030] In the example of Fig. 6, an area V that displays a surveillance video in real time and an area F that extracts a partial area including an abnormality occurrence area S from the surveillance video and displays it as a region of interest are arranged side by side on the display screen of the display device 7. The region of interest may be an image that is enlarged compared to the image of the surveillance video. As in the example of Fig. 6, by extracting and displaying the abnormality occurrence area S as a region of interest, the display device 7 (monitoring device 1) can draw the attention of a person watching the display device 7 (for example, a person monitoring the monitoring area) to the abnormality occurrence area S.

[0031] FIG. 7 shows another example of the display mode of the surveillance video and the abnormality occurrence area information on the display screen of the display device 7. In the example of FIG. 7, in a normal case where no abnormality occurs, the surveillance video is displayed. On the other hand, when an abnormality occurrence area is estimated by the estimation unit 13 (when an abnormality occurs), a region of interest including the abnormality occurrence area is extracted and superimposed on the surveillance video image. The position of the superimposed display of this region of interest is a part of the image of the surveillance video image that avoids the abnormality occurrence area. Even in such a case, the display device 7 (monitoring device 1) can draw the attention of a person watching the display device 7 (for example, a person monitoring the monitoring area) to the abnormality occurrence area S.

[0032] As another example of the display mode of the surveillance video and the abnormality occurrence area information on the display screen of the display device 7, a display mode as shown in Fig. 8 may be adopted. For example, in the example of Fig. 8, a figure (arrow mark) indicating the direction of the crowd where an abnormality was detected and a line indicating the outer edge of the abnormality occurrence area are superimposed on the surveillance video image. Even with such a display mode, attention can be drawn to the abnormality occurrence area S in the surveillance area. However, in this case, since a figure or line for attracting attention is superimposed on the abnormality occurrence area in the surveillance video, there is a risk that the image of the abnormality occurrence area itself will become difficult to see. Taking this into consideration, the shape, size, line thickness, etc. of the figure superimposed on the surveillance video are set.

[0033] The monitoring device 1 is configured as described above. Next, an example of an operation related to anomaly detection in the monitoring device 1 will be described with reference to Fig. 9. Fig. 9 is a flowchart illustrating an example of an operation related to anomaly detection in the monitoring device 1.

[0034] For example, first, the acquisition unit 11 acquires a surveillance video from the imaging device 5 (step 101 in FIG. 9). After that, the detection unit 12 executes anomaly detection processing in the surveillance area using the changes in the crowd and the direction of the crowd in the surveillance video detected by the crowd behavior analysis technology (step 102). Note that the changes in the crowd here refer to changes in the size (number of people and spread), location, congestion level (population density), etc. of the crowd.

[0035] On the other hand, the estimation unit 13 judges whether or not the detection unit 12 has detected an abnormality in the monitoring area through the abnormality detection process (step 103). If no abnormality is detected, the operations from step 101 onwards are repeatedly executed. On the other hand, if an abnormality is detected, the estimation unit 13 estimates an abnormality occurrence area in the monitoring area using information on the direction of the abnormal crowd detected in the monitoring video (step 104).

[0036] Thereafter, the output unit 14 outputs the abnormality occurrence region information (step 105). Then, based on information on a preset timing for ending monitoring and the presence or absence of a request to end monitoring, it is determined whether or not to end monitoring (step 106). If monitoring is to be ended, the operation related to abnormality detection is ended, but if monitoring is not to be ended, the operations from step 101 onwards are repeatedly executed.

[0037] The monitoring device 1, having the above-mentioned configuration, can achieve the following effects. That is, the monitoring device 1 detects an abnormality in a monitoring area by detecting an abnormality in a crowd in the monitoring area from a monitoring video, and further detects an abnormality occurrence area using the direction of the detected crowd. In this way, the monitoring device 1 can detect, for example, an abnormal state in a monitoring area with a flow of people (people passing) that is different from a normal state when viewed as a crowd, even though the behavior of each individual cannot be said to be abnormal, and can also detect an abnormality occurrence area. In other words, the monitoring device 1 can contribute to improving the performance of detecting an abnormality in a monitoring area with a flow of people (people passing).

[0038] Furthermore, the monitoring device 1 is provided with not only the standard criteria as criteria for judging the occurrence of an abnormality, but also a criterion for judging a low pedestrian flow state that is determined based on a pedestrian flow that is less than the pedestrian flow in the standard state. When the pedestrian flow detected in the monitoring area is less than the pedestrian flow in the standard state, the monitoring device 1 detects an abnormality in the crowd in the monitoring area using the criterion for judging the low pedestrian flow state. As a result, the monitoring device 1 executes an abnormality detection process for the monitoring area using criteria according to the amount of pedestrian flow, thereby further improving the accuracy of abnormality detection in the monitoring area.

[0039] Although crowd behavior analysis technology is given as an example of a method used by the detection unit 12 to detect an anomaly, the method used by the detection unit 12 to detect an anomaly is not limited to crowd behavior analysis technology. For example, the detection unit 12 may detect an anomaly in a monitoring area using the following method. That is, the detection unit 12 detects people from a monitoring video, and from this detection result, judges whether or not there is a crowd (group) in the monitoring area using population density and the like. Then, when the detection unit 12 judges that there is a crowd in the monitoring area, it estimates the size of the crowd (e.g., the number of people) and the direction of the crowd. Furthermore, the detection unit 12 judges whether or not the state of the crowd is abnormal using the direction of the crowd and the like. At the time of this judgment, if the crowd is larger than a preset size, any one or more of the number of people constituting the crowd, the population density of the crowd, and the area occupied by the crowd in the monitoring area may be further used to detect an anomaly in the crowd.

[0040] Furthermore, in the explanation of the first embodiment, an example is shown in which the above-mentioned standard judgment criteria and the judgment criteria of a low-people-flow state for judging whether the state of a crowd is abnormal are applied to the entire monitoring area. Alternatively, for example, the monitoring area may be divided into a plurality of regions, and the standard judgment criteria and the judgment criteria of a low-people-flow state may be set for each divided region. In this case, it is determined for each divided region which of the standard judgment criteria and the judgment criteria of a low-people-flow state will be used by a method similar to that described above.

[0041] Furthermore, when determining whether to use the standard judgment criteria or the criterion for judging a low-people flow state, factors that affect the size of the crowd (people flow), such as the time of day, weather, and the presence or absence of events in the monitoring area and its surroundings, may be taken into consideration. Furthermore, in determining the judgment criteria as described above, when the crowd is larger than a preset size, any one or more of the number of people constituting the crowd, the population density of the crowd, and the area occupied by the crowd in the monitoring area may also be used.

[0042] <Other embodiments> The present disclosure is not limited to the above-described first embodiment, and may be implemented in various ways. For example, in the first embodiment, an example is described in which the output destination to which the output unit 14 outputs the abnormality occurrence area information is the display device 7. In addition to the display device 7, the output unit 14 may also output the abnormality occurrence area information to, for example, a terminal device that is preset as an output destination. The terminal device is, for example, an information device equipped with a communication function, such as a personal computer (PC), a smartphone, or a tablet terminal device, and is operated, for example, by a monitor who monitors the monitoring area.

[0043] In the first embodiment, an example is shown in which the monitoring device 1 is separate from the photographing device 5. Alternatively, for example, the monitoring device 1 may be mounted on the photographing device 5. In this case, for example, the acquisition unit 11 of the monitoring device 1 directly acquires a monitoring video from a device functional part in the photographing device 5 that processes photographed images.

[0044] Furthermore, in the first embodiment, an example is shown in which the monitoring device 1 is separate from the display device 7. Alternatively, the monitoring device 1 may be configured to be equipped with the display device 7. Furthermore, a part of the functions of the monitoring device 1 described in the first embodiment may be distributed to one or both of the imaging device 5 and the display device 7. Furthermore, although one monitoring device 1 has the functions of the monitoring device described in the first embodiment, for example, multiple monitoring devices may have the functions of the monitoring device described in the first embodiment in a distributed manner, and the multiple monitoring devices may cooperate to execute the process of detecting anomalies in a monitoring area. Furthermore, a server device is an example of a computer device that functions as the monitoring device 1. The server device may be either a physical server or a cloud server.

[0045] Furthermore, in the first embodiment, the detection unit 12 detects an abnormality in the monitoring area by detecting a crowd in an abnormal state in the monitoring area. In addition to this, for example, when the number of people detected in the monitoring area is very small compared to the flow of people in a standard state (for example, when the number of people detected is a preset small number such as 5% or less of the number of people in the standard state), the detection unit 12 may perform an abnormality detection in the monitoring area by focusing on individuals in the monitoring area. The method of anomaly detection focusing on individuals in this case is not limited here, and a description thereof will be omitted.

[0046] Furthermore, in the first embodiment, an example is described in which the monitoring device 1 monitors one monitoring area, but the monitoring device 1 may monitor multiple monitoring areas in parallel. Furthermore, in the first embodiment, an example is shown in which the detection unit 12 sequentially processes monitoring videos acquired in real time to detect abnormalities in the monitoring area. Alternatively, for example, the detection unit 12 may read out the monitoring videos acquired from the image capture device 5 and stored in the storage device 20 at a preset timing, and execute the process of detecting abnormalities in the monitoring area using the read monitoring videos.

[0047] The monitoring device in the present disclosure may have a configuration as shown in FIG. 10. That is, in the example of FIG. 10, the monitoring device 30 is, for example, a computer device, and includes an acquisition unit 31, an estimation unit 32, and an output unit 33 as functional units realized by executing a computer program. The acquisition unit 31 acquires a video of a monitoring area as a monitoring video. An example of the acquisition unit 31 is the acquisition unit 11 described in the first embodiment. The estimation unit 32 estimates an abnormality occurrence area where an abnormality has occurred in the monitoring area by using information on the direction of the crowd detected from the monitoring video. An example of the estimation unit 32 is the estimation unit 13 described in the first embodiment. The output unit 33 outputs information (abnormality occurrence area information) representing an abnormality occurrence area in the monitoring area. An example of the output unit 33 is the output unit 14 described in the first embodiment.

[0048] An example of the operation of the monitoring device 30 shown in Fig. 10 will be described with reference to Fig. 11. Fig. 11 is a flowchart illustrating an example of the operation of the monitoring device 30.

[0049] For example, the acquisition unit 31 of the monitoring device 30 acquires a monitoring video (step 201). When an abnormality in a monitoring area is detected from the acquired monitoring video and information on the direction of a crowd detected in the monitoring area is received, the estimation unit 32 estimates an abnormality occurrence area in which the abnormality has occurred in the monitoring area using the information on the direction of the crowd (step 202). After that, the output unit 33 outputs the abnormality occurrence area information (step 203).

[0050] The monitoring device 30 has the configuration described above, and therefore can contribute to improving the performance of detecting abnormalities in a monitoring area where there is a flow of people (traffic).

[0051] A part or all of the above-described embodiments can be described as, but is not limited to, the following supplementary notes.

[0052] (Appendix 1) an acquisition unit that acquires a video captured in a surveillance area as a surveillance video; an estimation unit that estimates an abnormality occurrence region where an abnormality has occurred in a surveillance area by using information on the direction of the crowd detected from the surveillance video; an output unit that outputs information indicating an abnormality occurrence area in a monitoring area; A monitoring device comprising: (Appendix 2) The estimation unit estimates an abnormality occurrence area using a direction of the detected crowd in an abnormal staying state when a crowd in a predetermined abnormal staying state is detected in a monitoring area. 2. A monitoring device as described in claim 1. (Appendix 3) The output unit outputs the acquired monitoring video to a display device, and also outputs information representing the estimated abnormality occurrence area to the display device. 2. A monitoring device as described in claim 1. (Appendix 4) The system further includes a detection unit that detects an abnormality in the monitoring area using information on a change in the crowd and the direction of the crowd detected from the monitoring video. 2. A monitoring device as described in claim 1. (Appendix 5) The detection unit compares a given standard flow of people in the monitored area with the flow of people detected from the monitored video, and detects an abnormality occurring in the monitored area using the result of the comparison. 5. A monitoring device as described in claim 4. (Appendix 6) When the flow of people detected from the surveillance video is less than the flow of people in a standard state in the surveillance area, the detection unit detects the occurrence of an abnormality in the surveillance area by using a predetermined abnormality occurrence judgment criterion that is set based on a state in which the flow of people is less than the standard state. 6. A monitoring device as described in claim 5. (Appendix 7) The output unit causes the display device to display information representing the abnormality occurrence area superimposed on a monitoring video of the monitoring area. 4. A monitoring device as described in claim 3. (Appendix 8) The output unit causes the display device to display an area for displaying a surveillance video and an area for extracting and displaying a partial area including an abnormality occurrence area from the surveillance video side by side. 4. A monitoring device as described in claim 3. (Appendix 9) By computer, The video captured in the surveillance area is acquired as surveillance video. Using crowd direction information detected from surveillance video, anomaly occurrence areas in the surveillance area are estimated, Output information showing abnormality areas in the monitoring area Monitoring method. (Appendix 10) When estimating an abnormality occurrence area, if a crowd of people in a predetermined abnormally stagnant state is detected in a monitoring area, the direction of the detected crowd of people in the abnormally stagnant state is used to estimate the abnormality occurrence area. 9. The monitoring method described in Appendix 9. (Appendix 11) Regarding the output of information representing the abnormality occurrence area, the acquired monitoring video is output to a display device, and information representing the estimated abnormality occurrence area is also output to the display device. 9. The monitoring method described in Appendix 9. (Appendix 12) The computer also Detects abnormalities in a surveillance area using changes in crowds and information on the direction of the crowds detected from surveillance video. 9. The monitoring method described in Appendix 9. (Appendix 13) When detecting an abnormality in a monitored area, the flow of people in a standard state in advance in the monitored area is compared with the flow of people detected from the monitored video, and the occurrence of an abnormality in the monitored area is detected using the result of the comparison. 13. The monitoring method according to claim 12. (Appendix 14) When the flow of people detected from the surveillance video is less than the flow of people in a standard state in the surveillance area, a predetermined criterion for determining the occurrence of an abnormality is used to detect the occurrence of an abnormality in the surveillance area, the criterion being preset based on the state of a flow of people less than the standard state. 13. The monitoring method according to claim 13. (Appendix 15) The information representing the abnormality occurrence area is superimposed on the monitoring video of the monitoring area and displayed on the display device. 12. The monitoring method described in Appendix 11. (Appendix 16) An area for displaying a surveillance video and an area for extracting and displaying a partial area including an abnormality occurrence area from the surveillance video are displayed side by side on the display device. 12. The monitoring method described in Appendix 11. (Appendix 17) A process of acquiring a video of a monitoring area as a monitoring video; A process of estimating an abnormality occurrence area in a monitoring area using crowd direction information detected from a monitoring video; A process of outputting information indicating an abnormality occurring area in a monitoring area; A computer program for causing a computer to execute the above. (Appendix 18) When estimating the abnormality occurrence area, if a crowd of people in a predetermined abnormal staying state is detected in a monitoring area, the computer is caused to execute a process of estimating the abnormality occurrence area using the direction of the detected crowd of people in the abnormal staying state. 18. The computer program of claim 17. (Appendix 19) With regard to the output of information representing the abnormality occurrence area, a process of outputting the acquired monitoring video to a display device and outputting information representing the estimated abnormality occurrence area to the display device is executed by the computer. 18. The computer program of claim 17. (Appendix 20) The computer is further made to execute a process of detecting an abnormality occurring in the monitoring area using information on the change in the crowd and the direction of the crowd detected from the monitoring video. 18. The computer program of claim 17. (Appendix 21) When detecting an abnormality in a monitored area, a computer is caused to execute a process of comparing a pedestrian flow in the monitored area under a predetermined standard condition with a pedestrian flow detected from a monitored video, and detecting an abnormality in the monitored area using the result of the comparison. 21. The computer program of claim 20. (Appendix 22) When the flow of people detected from the surveillance video is less than the flow of people in a standard state in the surveillance area, a computer is caused to execute a process of detecting the occurrence of an abnormality in the surveillance area using a predetermined criterion for determining the occurrence of an abnormality that is set based on a state in which the flow of people is less than the standard state. 22. The computer program of claim 21. (Appendix 23) A computer is caused to execute a process of superimposing information representing an abnormality occurrence area on a monitoring video of a monitoring area and displaying the information on the display device. 20. The computer program of claim 19. (Appendix 24) A process is performed in which a region for displaying a surveillance video and a region for extracting and displaying a partial region including an abnormality occurrence region from the surveillance video are displayed side by side on the display device. 20. The computer program of claim 19.

[0053] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-mentioned embodiments. Various modifications that can be understood by a person skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, each embodiment can be appropriately combined with other embodiments. [Explanation of symbols]

[0054] 1,30 Monitoring equipment 11,31 Acquisition Department 12 Detection unit 13,32 Estimation part 14,33 Output section

Claims

1. an acquisition unit that acquires a video captured in a surveillance area as a surveillance video; an estimation unit that estimates an abnormality occurrence region where an abnormality has occurred in a surveillance area by using information on the direction of the crowd detected from the surveillance video; an output unit that outputs information indicating an abnormality occurrence area in a monitoring area; A monitoring device comprising:

2. The estimation unit estimates an abnormality occurrence area using a direction of the detected crowd in an abnormal staying state when a crowd in a predetermined abnormal staying state is detected in a monitoring area. The monitoring device according to claim 1.

3. The output unit outputs the acquired monitoring video to a display device, and also outputs information representing the estimated abnormality occurrence area to the display device. The monitoring device according to claim 1.

4. The system further includes a detection unit that detects an abnormality in the monitoring area using information on a change in the crowd and the direction of the crowd detected from the monitoring video. The monitoring device according to claim 1.

5. The detection unit compares the flow of people in a monitored area under a predetermined standard condition with the flow of people detected from the monitored video, and detects the occurrence of an abnormality in the monitored area using the result of the comparison.

5. The monitoring device according to claim 4.

6. When the flow of people detected from the surveillance video is less than the flow of people in a standard state in the surveillance area, the detection unit detects the occurrence of an abnormality in the surveillance area by using a predetermined abnormality occurrence judgment criterion that is set in advance based on a state in which the flow of people is less than the standard state. The monitoring device according to claim 5.

7. The output unit causes the display device to display information representing the abnormality occurrence area superimposed on a monitoring video of the monitoring area. The monitoring device according to claim 3.

8. The output unit causes the display device to display an area for displaying a surveillance video and an area for extracting and displaying a partial area including an abnormality occurrence area from the surveillance video side by side. The monitoring device according to claim 3.

9. By computer, The video captured in the surveillance area is acquired as surveillance video. Using crowd direction information detected from surveillance video, anomaly occurrence areas in the surveillance area are estimated, Output information showing abnormality areas in the monitoring area Monitoring method.

10. A process of acquiring a video of a monitoring area as a monitoring video; A process of estimating an abnormality occurrence area in a monitoring area using crowd direction information detected from a monitoring video; A process of outputting information indicating an abnormality occurring area in a monitoring area; A computer program for causing a computer to execute the above.

Images