JP4937016B2 - Monitoring device, monitoring method and program - Google Patents

Description

The present invention relates to a system for monitoring the intrusion of a moving object such as a person or a car into a monitoring area such as a factory site or a building.
In the following description, the moving body is assumed to be a person, but is not specified as a person.
In general, remote monitoring of an area is performed by a monitoring camera.
In recent years, surveillance cameras convert a surveillance video into a standard digital video format such as Motion JPEG or MPEG, and there are an increasing number of models capable of transmitting this to an IP network.
The present invention additionally uses information on an entrance / exit control system (Access Control System, hereinafter referred to as ACS) and an intrusion sensor for specifying a person for switching when displaying a video from a surveillance camera. It is related to a system in which a plurality of monitoring devices are linked on a network to recognize, collate and track with higher accuracy than before and effectively display them on a monitoring screen.
The present invention also relates to a technique for learning a moving route between areas in which the moving body can move.

For the purpose of monitoring security zones of important public facilities and areas such as ATMs of power plants, airports, stations, buildings, banks, and data centers (PC servers), network cameras are used for monitoring suspicious individuals, etc. The number of installed systems is increasing.
In a small system of several to a dozen or so units, the video of all cameras can be displayed on a video surveillance display that is divided into screens. This operation was common.
However, in a large-scale system with a large number of cameras installed, the system cost increases due to an increase in the burden on video surveillance personnel and the addition of recorders.

As countermeasures against this problem, there have been proposed a method for reducing the visual monitoring burden on the observer by automatically recognizing a monitoring target such as a person or a car in the camera image using a computer, or a method for linking the monitoring system. Yes.
A system that works with the camera, ACS, and sensor is also being developed with limitations.
JP 2003-324720 A JP 2006-146378 A JP 2003-324720 A JP 2005-12415 A JP 2006-180387 A

  The present invention improves the tracking accuracy of a moving body such as a suspicious person by linking a plurality of cameras, ACSs, and sensors in area security monitoring, and optimizes the image display to the monitoring staff in tracking. The goal is to reduce the burden on observers and reduce system costs.

  Device systems for monitoring areas include ACS, surveillance cameras, intrusion detection sensors, and the like. The characteristics and issues of each monitoring device are summarized as follows.

Generally, an ACS is a device / system for restricting access to an area, and is used to manage access to a closed area (room) such as an indoor server room that cannot be physically intruded. .
Usually, an authentication device such as a password, IC card, biometrics (fingerprint, glow, etc.) is installed at the entrance of the room, a person is identified, and only those authorized (permitted) to the area can pass.
Many recent ACSs can be managed by connecting to a network (IP protocol or the like). In the case of an authentication error, not only the gate cannot be passed, but also where the unauthorized intrusion is attempted is displayed on the area map on the screen of the ACS monitoring console.
However, since ACS is based on an authentication system, there is a problem that it is difficult to apply in an open monitoring space such as outdoors although the recognition of a moving object is high.

The surveillance camera can be applied even when the surveillance space is open.
However, there is a problem of the blind spot of the camera, and the identification accuracy of the subject (moving body) in the video is poor (Identification) only by the current image recognition technology of the camera. That is, it can be detected with relatively high accuracy that a moving body is shown in the video, but it is difficult to accurately and uniquely identify the moving body in the video.
In addition, when a large number of surveillance cameras are installed, a technology for providing monitoring support on the system side is required so that the surveillance target space can be efficiently monitored with limited supervisor resources.

As a sensor, for example, there is a wide area intrusion detection sensor system that detects using infrared rays or radio waves. This is for detecting an intruder on the site boundary and can detect an intrusion in a wider area than the surveillance camera, but it is difficult to identify an intruding mobile body due to the characteristics of the device. There is also the problem that certain false detections occur.
As other sensors, there are shake detection sensors (installed on a glass surface to detect breakage and penetration of glass), sensors using RFID, and the like.

There is no universal monitoring system.
In addition, there is a problem that each current monitoring system is independent and difficult to cooperate, or only limited cooperation is possible.
The limited cooperation is, for example, that the ACS and the surveillance camera are linked, and the camera video on the surveillance screen is connected to the surveillance camera (camera paired camera) in the vicinity where the ACS authentication occurs when the room authentication is performed by the ACS. For example, the display is switched.
In this system, the efficiency of the switching display of camera images is central. Further, in such a system, it is necessary to combine the ACS and the monitoring camera, but there are cases where it cannot always be provided in view of installation cost, and there is a problem in screen display in such a case.

Based on the requirements for the system described above, the known technology is examined.
The technique described in Patent Document 1 is a method of displaying a moving body image collated with overall feature information on a monitoring monitor or a monitoring person specifying a moving body to be tracked. It is necessary to determine whether or not the target moving body is within the monitoring space, which is insufficient for reducing the monitoring burden.

In Patent Document 2, in order to solve the technical problem described in Patent Document 1, the image analysis result between the cameras is collated, the tracking accuracy is improved, and the illegal access is detected and displayed in conjunction with the ACS system. Are listed.
However, there is a problem that the burden reduction is insufficient because a high-accuracy three-dimensional position database is required in advance and the introduction cost is high, and the tracking screen to be shown to the observer is only the switching of the camera video.

Patent Document 3 is a technique related to a monitoring system that links a plurality of cameras and tracks and monitors a moving object.
This is a method in which tracking information is directly exchanged between cameras and cameras via a network, and there is a problem that the cost of each camera device tends to increase and the cost of changing tracking logic is also high.

  Patent Document 4 relates to a technique for dividing a screen and synthesizing and displaying a current video and a video predicted to move next in tracking a moving object. However, there is a problem in tracking accuracy due to the cooperation of only the camera.

  Patent Document 5 describes a technique for monitoring an RFID sensor and a monitoring camera in conjunction with each other.

There are also the following problems.
Route (drawing) information is required to track a moving object, but there is a problem that the cost of inputting information such as connection topology information and physical route distance between each monitoring camera, ACS, and sensor on the route is high. is there.

  One of the main objects of the present invention is to solve the above-mentioned problems. A plurality of monitoring devices such as a monitoring camera, an ACS, and a sensor are linked on a network to enter a monitoring target space. It is possible to highly accurately recognize, collate, and track a person, display the video from the surveillance camera effectively on the monitoring screen, calculate the estimated arrival time of the intruder to each surveillance camera, and estimate the arrival time The main purpose is to display.

The monitoring device according to the present invention is:
A captured video receiver that receives captured video from a plurality of cameras arranged in the monitoring target space;
Estimated shooting that selects one or more cameras that are estimated to be able to shoot the moving body when the moving body enters the monitoring target space and the moving body reaches the shootable range in the near future. Possible camera selection part,
An estimated arrival time calculation unit that calculates, for each estimated image-capable camera, a time estimated that the moving body reaches the image-capable range of the estimated image-capable camera;
A display unit configured to output to the display device a captured image of the estimated shootable camera received by the captured image reception unit and an estimated arrival time of each estimated shootable camera calculated by the estimated arrival time calculation unit; It is characterized by that.

  According to the present invention, it is possible to select an estimated shootable camera that is estimated to be able to shoot a moving body in the near future, calculate an estimated arrival time of the moving body for each estimated shootable camera, and display the calculated estimated arrival time Therefore, when the inspector tracks the intruder, the burden on the inspector of guessing the intruder's movement route is eliminated, and the approximate time when the intruder appears in the monitoring video is notified to the inspector. Thus, the monitoring person can know the timing to watch each monitoring video, and the burden on the monitoring person in tracking the intruder can be reduced, and the tracking loss can be reduced.

Embodiment 1 FIG.
In the system according to the present embodiment, first, a plurality of cameras are installed in the monitoring target space, and means for extracting a moving body from the video is provided using an image recognition technique.
Background subtraction is well known as a method for detecting a moving object from a camera image. In the moving body detection process, the video may be sent to the server via the network and processed on the receiving server, or a detection mechanism may be incorporated in the camera. It is important to have detection means and notification methods. Note that it is possible to detect a moving object by means other than the background subtraction method.

  In addition, detection (tracking) of moving objects that have entered the monitored space includes detection of moving objects by image analysis of a monitoring camera (in a portion that is not a normal route) at the area boundary, detection by a sensor installed at the boundary, etc. Conceivable.

  Only the moving body recognition / tracking technique using the monitoring camera has a problem of accuracy, and thus the authentication accuracy and position accuracy of the target moving body are corrected by information such as ACS and sensor.

In addition, in order to reduce the monitoring burden on the surveillance staff, a screen in which the current video showing the target moving body and the adjacent camera video that can be calculated from the position information are combined and displayed on the same screen is displayed. There is an ACS and a sensor between the local point and the adjacent camera, thereby generating correction information of the moving body, correcting the positional accuracy of the moving body based on this information, and dividing / combining it on the screen.
The tracking image of the moving object can be divided and displayed by the adjacent camera video. For example, when the position information is supplemented by the ACS but the corresponding monitoring camera is not installed, the tracking image is moved to the observer using the ACS layout diagram. Alert the body position.

FIG. 1 is a diagram illustrating a system configuration example according to the present embodiment.
A plurality of surveillance cameras 101, an entry / exit management device 102 (ACS), and sensors 103 are installed via the network 100.
One or more monitoring cameras 101 are installed and monitored in a certain monitoring target space (hereinafter also referred to as monitoring space).
It is also possible to monitor the monitoring camera 101 by grouping the sensor 103 and the ACS 102 (grouping) (monitoring spaces 122 and 123). When grouped, the ACS 102, the sensor 103, and the monitoring camera 101 can be linked and monitored.
The grouping information is managed by the mobile monitoring device 104 connected to the network 100.
The monitoring target in the mobile monitoring device 104 is a mobile object such as a person or a car that enters the monitoring target space.

The network 100 will be described below on the assumption of an IP network.
However, the network 100 is not limited to an IP network, and it is sufficient that a means for transmitting video and information is secured. Communication may be wired or wireless (such as a wireless LAN).

The surveillance camera 101 captures an image of the surveillance space, encodes the captured image, and then transmits the image to the mobile object monitoring device 104 or the like via the network 100.
The mobile monitoring device 104 decodes the received video data and displays the video. The video encoding method is Motin-JPEG, MPEG2, MPEG4, H.264. There are various standard systems such as H.264, which are used.

The entry / exit management device 102 identifies the person by reading the ID information of the IC card with a card reader, for example, allows only the authorized person to enter the room, and notifies other devices of the entry / exit history etc. Has function.
For identification, in addition to an IC card, bio-authentication using a fingerprint or the like is known.

The sensor 103 has a function of detecting a moving body and notifying the location where the intrusion of the moving body is detected via the network. There are various types of sensors. For example, there is a contact sensor that detects opening and closing of a door.
In addition, there is an optical fiber detection sensor that detects an intrusion by specifying a position of bending, vibration, breaking, or pulling that occurs in a fence.
Furthermore, there are wide area intrusion detection sensors that detect changes in electromagnetic fields and detect intrusions of moving objects, and vibration sensors attached to glass. The main purpose of these sensors is to specify the position where the moving body is estimated to exist.
The position information may be stored in the sensor 103 or may be stored in the moving object monitoring device 104.
When the sensor 103 detects an event, the event is notified to the moving body monitoring device 104. If the notified event does not have position information, the moving body monitoring device 104 complements the position by the monitoring database 114 from the ID for identifying the sensor.
Further, the sensor 103 may be an RF-ID reader, and in this case, the moving body can be identified from the read card ID.

The mobile object monitoring device 104 is connected to the network 100 as described above.
The moving object monitoring apparatus 104 has a function of tracking the moving object by controlling the linkage of the monitoring camera 101, the ACS 102, and the sensor 103.
A video recording device 105, a monitoring display 106, and a camera changeover switch 107 can be connected to the moving object monitoring device 104.
The video recording device 105, the monitoring display 106, and the camera changeover switch 107 are directly connected to the network 100 (LAN), and may be connected to the moving object monitoring device 104 via the network.

The video recording device 105 records (accumulates) video captured and transmitted by the monitoring camera 101. Usually, many are written and stored in a hard disk device (internal or external).
Although not shown in FIG. 1, when this system is connected to an external information system and an unauthorized intruder is detected, it is possible to make a call to an e-mail, a mobile phone or the like.

The monitoring display 106 displays the received video of the monitoring camera on a screen divided into N, and enables video monitoring of a plurality of monitoring spaces on the same screen.
The monitoring display 106 is an example of a display device.

The camera changeover switch 107 is a device for the supervisor to manually switch the camera image displayed on the monitoring display 106.
Some are connected via the mobile monitoring device 104, some are directly connected to the video recording device 105 to be controlled, and others are connected via the network 100.
The camera changeover switch 107 is a monitoring console (input device) used by a monitor.

Next, the internal configuration of the moving object monitoring apparatus 104 will be described.
The camera image receiving unit 120 receives captured images captured by the plurality of monitoring cameras 101 arranged in the monitoring target space via the network 100.
The camera video receiving unit 120 is an example of a captured video receiving unit.

The camera image moving body detection unit 110 has a function of detecting a moving body from an image in a specific camera.
Image recognition technology is used to detect moving objects. Specifically, a moving object is detected by comparing each video frame in time series with a difference. Existing means are used to identify and extract a person from a video frame.
The camera moving body detection unit 110 is an example of a moving body identification unit.

The inter-camera movement path calculation unit 111 is a function that calculates the path, distance, and time expected to move next from the current position of the moving body using the inter-camera path data in the monitoring database 114.
When the moving body enters the monitoring target space, the inter-camera movement path calculation unit 111 can reach the shootable range from the camera that is currently shooting the moving body and can shoot the moving body. One or more cameras to be estimated are selected as cameras that can be estimated. The imageable range is a spatial range around the installation position of the monitoring camera that can be captured (captured) by the monitoring camera.
Further, the inter-camera movement path calculation unit 111 calculates, for each estimated photographable camera, the estimated time when the moving body reaches the photographing possible range of the estimated photographing capable camera. When calculating the estimated arrival time, the inter-camera movement path calculation unit 111 estimates each estimated camera that can be photographed using the average moving time for the moving object attribute that matches the attribute of the moving object that has entered the monitoring target space. The arrival time is calculated.
The inter-camera movement path calculation unit 111 also measures the actual travel time required when the moving body that has entered the monitoring target space moves from the camera being photographed to another camera, and from the camera being photographed to each estimated photographable camera. Compare the average travel time for each mobile object attribute when moving between the two, and determine the mobile object attribute that approximates the actual travel time of the mobile object that has entered the monitored space. The estimated arrival time can also be calculated for each estimated camera capable of being photographed using the average moving time for the body attribute.
The inter-camera movement path calculation unit 111 is an example of an estimated photographing capable camera selection unit and an estimated arrival time calculation unit.

  The inter-camera moving body matching degree calculation unit 112 has a function of tracking a moving body by comparing the past recorded video of the moving body being tracked with the current moving body video of each camera and calculating the degree of matching. The inter-camera moving body matching degree calculation unit 112 extracts the feature amount of the moving body in each video. Compare extracted feature values and track those with similar extracted feature values.

  The ACS moving body detection unit 113 is a function that receives an ACS 102 entry / exit event, identifies a moving body that passes through the ACS 102, and calculates a position.

The sensor moving body detection unit 117 has a function of receiving a sensor event and calculating the position of the moving body.
For example, when the sensor moving body detection unit 117 receives a “door open” event from a contact sensor, the position in the monitoring database 114 is based on the sensor ID (network address, contact I / F number, etc.) included in the event as a key. The position where the door is opened can be specified using the information data.

  The position management unit 115 is a function for managing reference and registration of data related to position management, such as layout data in the monitoring database 114 and route data between cameras.

The video composition / display switching unit 116 superimposes the monitoring character information on the camera video, displays the video in the divided display area, and displays it according to the input of the monitor from the display control function or the camera selector switch 107. This is a function for switching camera images.
The video composition / display switching unit 116, for example, the captured video of the estimated shootable camera received by the camera video receiving unit 120, the estimated arrival time for each estimated shootable camera calculated by the inter-camera movement path calculation unit 111, and the like. Can be output to the monitoring display 106.
The video composition / display switching unit 116 is an example of a display unit.

The inter-camera path learning unit 118 generates an adjacent state of the camera in advance by setting a registered user as a sample moving body for generating path data, and averages depending on attributes (male, female, etc.) of the sample moving body This is a function for learning the travel time and generating the route data in the monitoring database 114 semi-automatically.
The inter-camera route learning unit 118 is an example of an average travel time calculation unit.

The behavior analysis intrusion determination unit 119 is a function that analyzes the behavior of a moving object from information such as camera images, ACS, and sensors to determine intrusion.
The behavior analysis intrusion determination unit 119 is a function for determining a behavior abnormality.
For example, in the authentication by ACS, this is a function for determining and notifying a behavior different from normal, for example, when a gate open failure with the same card (ID) continues N times, it is regarded as abnormal. The condition for behavior determination can be set.
The behavior analysis intrusion determination unit 119 is a function for performing intrusion determination. For example, when a moving body that does not normally pass by or moves over the fence on the outer periphery of the site is detected, it is determined whether it is a person or an animal such as a cat from the size of the moving body, and if it is a person, it is determined to be an intrusion. It has a function.
Conditions for behavior analysis and intrusion determination are also included, such as determining that a mobile object (person) is invaded when two or more moving objects (persons) are detected by the moving object detection unit 110 in the camera image during ACS passage.
If it is determined that the intrusion has occurred, the monitoring video is preferentially displayed on the monitoring display 106 to notify the monitoring person.

The monitoring database 114 stores information that allows each unit of the moving object monitoring apparatus 104 to function.
The monitoring database 114 stores information to be described later. For example, the monitoring database 114 stores average moving time information indicating the average moving time for each moving object attribute when moving between cameras.
The monitoring database 114 is an example of an average travel time information storage unit.

Here, information managed by the monitoring database 114 will be described.
For example, the monitoring database 114 manages the following information.
A) Setting information such as the network address of the camera, ACS (card reader, etc.), sensor (door open / close sensor).
B) Physical installation position (drawing data) data of the camera, ACS (card reader, etc.), sensor (door open / close sensor).
C) Camera, ACS, sensor grouping data.
D) Route data between cameras.
The route data is held as a camera adjacent to the camera, and the distance between the two cameras, the average moving time, and the like are also held.
E) Behavior determination data for determining fraudulent behavior F) User database such as user ID and traffic authority information registered in ACS.
Users have employee data and registered visitor data.
G) A database of divided display sections in the camera changeover switch and the monitor display.

FIG. 2 shows a physical installation example of a monitoring camera or the like in the monitoring target space.
FIG. 2 is an example of a system that monitors a certain range such as a factory site in cooperation with an outer peripheral sensor, a monitoring camera, and an ACS.
For example, there is a sensor for specifying the position of vibration, breaking, and pulling of the fence as an outer periphery sensor. These sensors have a mechanism for estimating (specifying within a certain range) the position, and the estimated position can be notified to the mobile monitoring apparatus 104 in FIG.
As shown in FIG. 2, some surveillance cameras, sensors, and ACS (card reader or the like) are installed in the vicinity. Usually, it installs so that the moving body (person) who penetrate | invaded the site and passed ACS can be caught with a surveillance camera.
In FIG. 2, C1 to C7 are outdoor surveillance cameras. In FIG. 2, C01, C02, and C03 are outdoor or indoor surveillance cameras installed at the entrance of the building. Broken lines around C1 to C7 indicate the monitorable space of each camera. The monitorable space is a shootable range of each camera. In FIG. 2, C01, C02, and C03 do not show a monitorable space that is an imageable range, but C01, C02, and C03 also have a monitorable space.
In areas where access restrictions are required, an entry / exit management device (ACS) is installed, and when strict security measures are not required, door sensors and surveillance cameras may be installed.
When the ACS or sensor detects the opening / closing of the door, the event is notified to the moving body monitoring device 104, and the monitoring video can be preferentially displayed in conjunction with the event.
Recently, there is also a network camera that can detect a moving object by performing video analysis inside the camera and notify the monitoring apparatus of the moving object detection event even with the monitoring camera alone.
In FIG. 2, cameras C11 and C12 represent surveillance cameras installed at doors on arbitrary floors of the building. Although C11 and C12 in FIG. 2 represent only one floor (first floor), in reality, entry and exit of security-restricted places such as server rooms in the entrances and rooms of each floor such as C21 and C22 are ACS. The video is monitored with a surveillance camera. In FIG. 2, C11 and C12 do not show a monitorable space that is an imageable range, but C11 and C12 also have a monitorable space.

FIG. 3 shows a video display flow of an estimated intrusion position at the time of intrusion detection (intrusion detection phase).
In parallel with the processing of FIG. 3, the video of each monitoring camera 101 is transmitted to the mobile monitoring device 104 via the network 100, and the camera video receiving unit 120 always receives the video from each monitoring camera 101. (Shooting video reception step).
In FIG. 3, for example, when a moving body passing through the perimeter sensor at the site boundary is detected (S <b> 301), the intrusion estimated position is notified to the mobile body monitoring device 104 through the outer circumference sensor device having a communication mechanism.
In the moving body monitoring device 104, the sensor moving body detection unit 117 receives a message from the outer circumference sensor, and acquires intrusion position information indicating an estimated intrusion position from the notified message (S302).
If the intrusion position information is not included in the message, the sensor installation position information in the monitoring database 114 may be searched and specified using the ID for specifying the notification source sensor as a key.
In the moving object monitoring apparatus 104, the position management unit 115 searches for the installation positions of the monitoring cameras in order from the estimated intrusion position as a key and lists the monitoring camera candidates.
Then, the video composition / display switching unit 116 displays the video of the camera arranged in the vicinity of the estimated intrusion position among the video images of the plurality of monitoring cameras received by the camera video receiving unit 120. Show as many as you can (preset).
For example, in FIG. 2, when the outer circumference sensor detects the “estimated intrusion position” in the figure, the surveillance cameras near the entry point are C4, C5, C3, C6,.
If there are three video display settings at the time of intrusion detection, C4, C5, and C3 videos are displayed in the display area of the monitor display.

The priority of display can be changed by the moving image analysis technology of the surveillance camera, in addition to the method of displaying priority from the above-mentioned ones.
For example, priority is given to those in which a moving body is reflected in the video by the moving body detection unit 110 in the camera video. For example, a background subtraction method is known as a specific technique for moving object detection. The background subtraction method saves video data in a state where there is no moving object as the background, and obtains the difference between the video data showing the moving object and the luminance value for each pixel. This is a method for obtaining the area of the moving body. Then, by using information such as the size of the person, a method such as dividing the moving body area into individual person areas is adopted. In the present embodiment, the camera image moving body detection unit 110 uses these known techniques.

In the intrusion detection phase, it is important to detect the moving object based on the estimated position and present it to the monitor.
Therefore, normally, the camera that detects the moving body is set to display with priority. The display control on the monitoring display is set as a display switching rule for the video composition / display switching unit 116 in the moving object monitoring device 104.
At the site boundary, the sensor may react due to the invasion of animals such as cats.
For this reason, the behavior analysis intrusion determination unit 119 of the moving object monitoring device 104 can reduce false detection by providing a certain threshold value for the value read by the sensor.
In addition, it is possible to provide a mechanism for notifying that there is a possibility of erroneous detection in the case of a moving body other than a person based on the size determination of the moving body by the above-described in-camera moving body detection unit 110.

Furthermore, by the following mechanism, intruders can be estimated not only from the current camera image but also from past images.
The camera video is received by the camera video receiver 120 and is always recorded in the video recording device 105 (recorder).
The intrusion position and time can be known from the sensor notification.
Therefore, the video taken after the sensor notification is analyzed, and the video from the camera located in the vicinity of the sensor is analyzed, and the past video showing the moving object is played back from the recording device, and the video is displayed in the designated area of the monitor display. Can be displayed.

  When the monitor finds a suspicious person in the video displayed by these means, the camera switch 107 is used to select the video (with the tracking start video) and simultaneously switch to the tracking monitoring mode.

Next, a video display processing flow at the time of tracking monitoring (tracking monitoring mode) will be described with reference to FIG.
First, the monitoring person operates the camera changeover switch 107 functioning as a monitoring console, and sets the selected video as a reference video for tracking (S401).
The reference video is saved at least during subsequent processing. Also, the moving body of the reference video is cut out and used for verification of subsequent processing.

  Then, the inter-camera movement path calculation unit 111 sets the camera of the video selected as the reference video by the monitor to the current (current) (S402).

Next, the inter-camera movement path calculation unit 111 searches for adjacent cameras to which the moving body can move from the current camera using the position information of the current camera (or sensor or the like) selected by the monitor as a key (S403) (S403) Estimated photographable camera selection step).
In other words, the inter-camera movement path calculation unit 111 is adjacent to the current camera that is shooting the moving body (the camera that is shooting), and the moving body is estimated to be capable of shooting the moving body in the near future. Select one or more cameras (cameras that can be estimated).

Next, the inter-camera movement path calculation unit 111 analyzes the current camera image and attempts to complement the attribute information of the moving object (intruder / suspicious person) (S404).
For example, the height, gender, age, and identification ID are registered in the database from the video of the person cut out by the moving body detection unit 110 in the camera video. If automatic analysis can be performed by image analysis technology, try to supplement using these means. However, since there are many cases that cannot be automatically complemented only by image analysis, the supervisor may input supplementary information from the camera changeover switch 107 that is a monitoring console.
These pieces of attribute information are used in the process for estimating the arrival time at the adjacent camera, which is a subsequent process.
Note that the complementary information of the attribute such as the intruder is for improving the movement estimation accuracy, and can be skipped instead of the essential processing.

  In FIG. 4, the intruder / suspicious person's attribute information is complemented after searching for the adjacent camera. However, after the intruder / suspicious person's attribute information is complemented, the neighboring camera is searched. Alternatively, these two processes may be performed in parallel.

Next, from the number of adjacent cameras searched in step S403 and the system setting value, the video composition / display switching unit 116 determines the number of divided screens and the video display position (S405).
For example, in the case of FIG. 2, when C4 is selected as the current camera, C03 and C02 are candidates as neighboring cameras in addition to C3 and C5. The number of divided screens is determined based on the number of adjacent cameras and the screen division setting value.

Next, the camera moving body detection unit 110 calculates the estimated arrival time of the moving object for each adjacent camera (S406) (estimated arrival time calculation step).
The estimated arrival time is a time at which the moving body reaches each imageable range of the adjacent camera, and it is predicted that the moving body can be captured by the adjacent camera.
As shown in 601 and 602 of FIG. 6, the estimated arrival time is expressed for each camera, for example, in the form that the mobile body arrives after 1 minute 30 seconds. Further, as another expression format, it may be expressed in the form of arriving at 12:10:30.
A specific method for calculating the estimated arrival time will be described later.

Next, the video composition / display switching unit 116 adds the attribute information (camera number / installation position) and estimated arrival time of each camera as additional information of the decoded video. The attribute information of each camera is added with the result of searching the monitoring database 114 (S407).
These pieces of information may be displayed as synthesized in a frame, or a character information display area may be provided in each divided screen of the monitoring display 106 to display these additional information separately from the video.

After the above processing, the video composition / display switching unit 116 outputs the video from the monitoring camera, the attribute information of the camera, and the estimated arrival time to the monitoring display 106 monitored by the monitor, and the monitoring display 106 tracks. The video, camera attribute information, and estimated arrival time are displayed (S408) (display step).
When the monitoring person confirms that the person being monitored has moved in the image of the adjacent camera, the monitoring person selects the camera image of the movement destination.
If the adjacent video is selected (YES in S409), the selected camera is newly set as the current camera, and the same processing as described above is performed.

If the monitor issues a tracking monitoring end command from the camera changeover switch 107 as the console (YES in S410), the tracking monitoring is ended and the normal monitoring mode (for example, rotation display of each camera image) is thereafter performed. To do.
If there is no monitoring end command (NO in S410), the monitoring video display is updated.
The timing for updating the monitoring video is updated for each video frame in a method of combining character information in an image. If the character information is displayed in a different area from the image information, the character information is usually updated by reducing the average moving time by the update period in a set update period (for example, every second).

FIG. 6 shows a display example of the tracking monitoring screen on the monitoring display 106.
Reference numeral 601 in FIG. 6 displays a captured image of the current camera in which the moving object is currently displayed and a captured image of an adjacent camera adjacent to the current camera. Also, the attribute information of each adjacent camera (for example, the camera number of the camera 5C and the like, the installation position such as the outdoor northeast) and the estimated arrival time of the adjacent camera (for example, after 1:50) are displayed together with the camera video.
In addition, as shown by 602 in FIG. 6, when the device adjacent to the current camera is an ACS or a sensor, a screen displaying the estimated position of the moving object is displayed on the layout diagram in addition to the divided display of the camera image. Also good.

Next, a method for calculating the estimated arrival time of the moving body at the adjacent camera will be described.
The time estimation is performed by the inter-camera movement route calculation unit 111 using information in the route information management database in the monitoring database 114.
FIG. 5 shows an outline of the route information management database.
The route information management database is information indicating the average moving time for each moving object attribute when moving between cameras, and is an example of average moving time information.

In the monitoring database 114, as the route information, basically, only adjacent information between cameras is stored.
In particular, indoor and outdoor cameras are usually installed together with ACS and sensors, and even if the camera is installed alone, it is installed so that no blind spot in the monitoring space is generated with respect to the passage. Basically, it is only necessary to store the information between cameras for the adjacent cameras.
However, because outdoor cameras have a large space to be monitored, blind spots are likely to occur, and the images are also subject to environmental influences such as weather and time. The possibility of monitoring omission increases.
Therefore, the information between two cameras is also stored as information between two points even if the cameras are not adjacent to each other.

The moving time between two points varies greatly depending on the attribute of the moving object, and the method of simply dividing the distance between the two points by the average speed of the moving object to obtain the estimated arrival time generally has poor accuracy.
For example, the travel time to reach the destination differs between employees who are familiar with the positional relationship and first-time visitors from outside. In addition, the arrival time differs even when moving on foot or on foot.
Therefore, in the monitoring database 114 according to the present embodiment, as shown in FIG. 5, the average moving time between two points is managed for each attribute of the moving object.
For example, in the table of FIG. 5, the average travel time A is the average travel time when an external first-time visitor / male who is not familiar with the site travels on foot. The average travel time B is the average travel time when employees / men who are familiar with the site travel on foot. The average moving time C is an average moving time when employees / men who are familiar with the site move quickly.
Thus, in the route information management database according to the present embodiment, the average moving time is managed based on the attributes of the moving object. If the attribute of the moving object is unknown, the average moving time A is used in advance.
In this route information management database, the average moving time for each moving object attribute is important, and the distance between two points is not essential, but it is collected as much as possible, and the information of the average moving time is not recorded Use. Information about whether the cameras are adjacent to each other may also be held.
Even if the movement is between the same two points, the movement time may be different if the movement is in the opposite direction. For example, first-time visitors are likely to be misidentified in the positional relationship between going and returning, and often differ. Also, the travel time may be different even when going up and down stairs. Therefore, the adjacent data is held so that the reverse movement can be distinguished.

Assuming that the database as described above is maintained, the calculation of the estimated arrival time in FIG. 4 can be easily derived as follows.
Hereinafter, a procedure in which the inter-camera movement path calculation unit 111 extracts the adjacent cameras of the current camera and derives the estimated arrival time of each adjacent camera will be described.
The inter-camera movement path calculation unit 111 selects a row in the table of FIG. 5 using the identification ID of the current camera that is the transmission source of the video selected by the monitor as a key.
For example, when the video of the camera C4 is now selected, the inter-camera movement path calculation unit 111 selects the row in which the start camera is C4 in the table of FIG. Since C4 is an outdoor camera, it is generally better to extract even if the adjacency flag is false (that is, when there is no adjacent relationship when the camera is set as a node in the graph and placed on the path graph).
Next, the inter-camera movement path calculation unit 111 extracts attribute information analyzed from a moving object in the camera image, or extracted from the attribute information input from the console by the observer viewing the image. The average moving time that is closest to the attribute of the moving object is selected. Thereby, the estimated movement time between two points is obtained.
After that, based on this information, the inter-camera movement path calculation unit 111 sequentially subtracts the elapsed time from the average movement time at the timing of updating the surveillance camera video in which the moving object is reflected, and obtains the estimated arrival time. The display switching unit 116 displays the estimated arrival time on the monitoring display 106.

About the average movement time employ | adopted as the average movement time for every attribute, it can also select as follows.
First, when the attribute information is unknown, the inter-camera movement route calculation unit 111 estimates using the average average movement time (default value) in the first route estimation.
When the moving body moves from the source monitoring camera to the destination monitoring camera, that is, when the actual moving time between two points (actually measured moving time) is determined, a plurality of averages between the two points are determined. The average travel time closest to the actually measured travel time is retrieved from the travel times, and the mobile object attribute is set as a new estimated attribute of the mobile object, which is used for the next travel time estimation of the destination.
The actual movement time between two points by the moving body may be measured manually by a monitoring person or may be measured using the method described in the second embodiment. Moreover, you may measure by another method.

As described above, the monitoring system including the mobile monitoring device according to the present embodiment detects and tracks unauthorized intrusion in security zones such as airports, factories, etc., office buildings, financial (night) unmanned stores, and the like. It can be used as a monitoring system.
And according to the present embodiment, the tracking accuracy for the moving body in the monitoring target area is improved, and the burden on the monitoring person is reduced by performing video display suitable for tracking the moving body of interest, As a result, tracking loss can be reduced.

As described above, in the present embodiment, in a system in which a plurality of cameras, entrance / exit management devices, and sensors are installed in a space to be monitored, the camera, the entrance / exit management device, and the means for detecting / identifying the moving body by the sensors A monitoring system including a means for calculating a movement route between cameras of a moving body, a monitoring database, and a display device for displaying the moving target image and state has been described.
In this embodiment, when tracking a moving object, the moving camera currently displays a monitoring camera image and a monitoring camera image of an adjacent camera that may travel, and displays a moving destination adjacent camera. A monitoring system that supports which of the adjacent images should be noted by combining and displaying the screen by means such as estimated arrival time (how much it will arrive later) and highlighting of divided frames has been described.

Embodiment 2. FIG.
When the processing shown in the first embodiment is assumed, a means for easily and automatically generating the average movement time for each attribute in the route information management database of FIG. 5 is desired.
Therefore, in the present embodiment, a procedure for generating an average moving time between two cameras will be described.

In the moving body monitoring apparatus 104 shown in FIG. 1, the moving body detecting unit 110 in the camera image analyzes a captured image from the first camera and selects a specific sample moving body from the captured images from the first monitoring camera. And analyzing a captured video from the second camera to identify a specific specimen moving body from the captured video from the second surveillance camera.
Further, the inter-camera path learning unit 118 confirms that the moving body detection unit 110 in the camera image has moved away from the shootable range of the first camera in the captured image from the first camera. The difference between the identified time and the time at which the specific sample moving body is identified as being within the shootable range of the second camera from the captured image from the second camera is calculated from the first camera to the second. Calculated as the average travel time to the camera.
In calculating the average moving time, the inter-camera path learning unit 118 makes the sample moving body a person who understands the moving path of the monitoring target space and a person who does not understand the moving path, so that the inter-camera path learning unit 118 The average travel time when a person who understands the travel route walks, the average travel time when the person who understands the travel route of the monitored space travels (when running), the travel route of the monitored space The average moving time can be calculated for each attribute of the moving body such as the average moving time when a person who does not understand (outsider) walks.
The monitoring database 114 stores the average moving time for each attribute of the moving object calculated by the inter-camera path learning unit 118, that is, the average moving time when a person who understands the moving path of the monitoring target space walks. The average movement time when a person who understands the movement path of the monitoring target space travels, the average movement time when a person who does not understand the movement path of the monitoring target space walks, and the like are stored.

Next, the operation at the time of deriving the average travel time in the present embodiment will be described.
In the present embodiment, as a method for obtaining the average moving time between two cameras when the camera is often installed alone, such as an outdoor camera, the method using a specimen moving body is performed as follows.
First, prepare a specimen moving body that can be easily identified by image analysis. For example, a specimen moving body (registered user) wears a bib (a symbol such as a number or a circle) that displays a symbol with high recognition accuracy by image analysis.
In the moving object monitoring device 104, the camera of two points (start point, end point) to be measured is selected, the attribute of the sample moving object (employee or outsider, male or female, walking or running, etc.), for identification Specify the image and start measurement.
In this state, the specimen moving body moves from the start point to the end point.
The moving body to be a specimen moves so as to be photographed by the start point camera and the end point camera.
In the moving body monitoring apparatus 104, the camera image receiving unit 120 receives the images of the start point camera and the end point camera, and the in-camera image moving body detection unit 110 analyzes the start point camera image and the end point camera image. . The in-camera moving body detection unit 110 analyzes whether or not the specimen moving body has disappeared from the video of the starting point camera, that is, whether or not the specimen moving body has fallen out of the shooting range of the starting point camera. It is determined whether or not a specimen moving body has appeared in the video of the camera, that is, whether or not the specimen moving body has entered the shootable range of the end point camera.
The inter-camera path learning unit 118 starts time measurement from the time when the sample moving body deviates from the shootable range of the starting camera, and completes measurement when the sample moving body enters the shootable range of the end camera. Find the time taken to move between two points. The time taken for the movement between the two points is defined as the average movement time.
Since this process does not require real-time characteristics, data can also be generated by analyzing an image once stored in the video recording apparatus.

Next, how to determine the travel time when an ACS device (or sensor) is installed in combination with a camera will be described. In ACS, the time of passing through each point is recorded for each ID (for each person).
In such a case, the average moving time of the monitoring camera between two points can be obtained by using the passage time of the ACS (or sensor) door.
For example, when employee A moves (enters office) from C1 to C01 shown in FIG. 2, the IC card reader is read into ACS-1, and then read into ACS-02 for authentication / authorization processing, and the door is opened. pass.
Since the traffic records are recorded as ACS logs, it is possible to sort by time for each employee and obtain the average travel time between cameras paired with each ACS.
After obtaining the average movement distance between the points of each employee, the average movement time of each attribute between the two points can be obtained by classifying the movement average of each employee for each attribute and averaging this.
In addition, the average travel time of a moving object (person) who is not familiar with the site, such as a visitor, is usually lent out for visiting cards, so the logs in the ACS are sorted by time of visitor lending card ID. However, the travel time can be obtained by calculating from the passing time between the two points.
That is, in the present embodiment, the inter-camera path learning unit 118 includes the time when a specific moving object passes through the first entrance / exit management device or the first sensor arranged in the vicinity of the first camera, Information (log) indicating the second entry / exit management device or the second sensor disposed near the second camera and the time when the specific moving body passes through the first entry / exit management device or The difference between the time when the specific moving body passes the first sensor and the time when the specific moving body passes the second entrance / exit management device or the second sensor is determined from the first camera to the second camera. Calculated as the average travel time until.

In addition, as described above, the installation position of each surveillance camera is often set when the camera is installed.
For example, if it is outdoors, the installation position can be accurately measured by using a GPS (Global Positioning System) or the like and set in the monitoring database.
In addition, for indoors, installation location information can be set and stored from the camera installation location on the design drawing of the building. It is also possible to obtain the distance between the two points from these position information, obtain the average moving speed of the moving body from the distance between the two points and the image analysis of the surveillance camera, etc., and estimate the destination and arrival time.

As described above, in order to calculate a tracking route, each monitoring camera, ACS, inter-sensor connection information and physical route information (database) on the moving route are required. It takes time and effort to manually input from this information, and the introduction cost increases.
For this reason, the mobile body monitoring apparatus according to the present embodiment uses a mobile body that can easily identify an ID (in the case of image recognition, for example, a person with a high image recognition rate or a person who has clothes) on the route. It has a function of moving two points (for example, adjacent cameras) and making adjacent information, moving average time, and the like into a DB.
And the data regarding a movement path | route can be learned by a sample moving body, and the input cost of drawing information can be held down.

As described above, in the present embodiment, the automatic generation processing of the average movement time of the route database has been described.
Specifically, in the generation of moving average time data, the adjacent state between cameras and the average moving time are automatically generated using a sample moving body with high analysis accuracy of passage logs and image analysis, and the arrival time estimation is made efficient. The processing method has been described.

Finally, a hardware configuration example of the moving object monitoring apparatus 104 shown in the first and second embodiments will be described.
FIG. 7 is a diagram illustrating an example of hardware resources of the mobile monitoring apparatus 104 shown in the first and second embodiments.
Note that the configuration of FIG. 7 is merely an example of the hardware configuration of the mobile monitoring device 104, and the hardware configuration of the mobile monitoring device 104 is not limited to the configuration shown in FIG. There may be.

In FIG. 7, the mobile monitoring device 104 includes a CPU 911 (also referred to as a central processing unit, a central processing unit, a processing unit, a processing unit, a microprocessor, a microcomputer, and a processor) that executes a program. The CPU 911 is connected to, for example, a ROM (Read Only Memory) 913, a RAM (Random Access Memory) 914, a communication board 915, a display device 901, a keyboard 902, a mouse 903, and a magnetic disk device 920 via a bus 912. Control hardware devices. Further, the CPU 911 may be connected to an FDD 904 (Flexible Disk Drive), a compact disk device 905 (CDD), a printer device 906, and a scanner device 907. Further, instead of the magnetic disk device 920, a storage device such as an optical disk device or a memory card read / write device may be used.
The RAM 914 is an example of a volatile memory. The storage media of the ROM 913, the FDD 904, the CDD 905, and the magnetic disk device 920 are an example of a nonvolatile memory. These are examples of the storage device.
The communication board 915, the keyboard 902, the mouse 302, the scanner device 907, the FDD 904, and the like are examples of input devices.
The communication board 915, the display device 901, the printer device 906, and the like are examples of output devices.
The monitoring database 114 and the video recording device 105 are realized by a magnetic disk device 920, for example.
The camera changeover switch 107 is realized by a keyboard 902 and a mouse 302, for example.
The monitoring display 106 is realized by the display device 901, for example.

As shown in FIG. 1, the communication board 915 is connected to a network. For example, the communication board 915 may be connected to a LAN (local area network), the Internet, a WAN (wide area network), or the like.
The magnetic disk device 920 stores an operating system 921 (OS), a window system 922, a program group 923, and a file group 924. The programs in the program group 923 are executed by the CPU 911, the operating system 921, and the window system 922.

The program group 923 stores programs for executing the functions described as “˜units” in the description of the first and second embodiments.
The program is read and executed by the CPU 911.

In the file group 924, in the description of the first and second embodiments, “detection of”, “determination of”, “calculation of”, “comparison of”, “search of”, “evaluation of”. ”,“ Update of ”,“ Setting of ”,“ Registration of ”, etc. Information, data, signal values, variable values, and parameters indicating the results of the processing are described as“ ˜file ”or“ ˜ It is stored as each item of “Database”.
The “˜file” and “˜database” are stored in a recording medium such as a disk or a memory. Information, data, signal values, variable values, and parameters stored in a storage medium such as a disk or memory are read out to the main memory or cache memory by the CPU 911 via a read / write circuit, and extracted, searched, referenced, compared, Used for CPU operations such as calculation, calculation, processing, editing, output, printing, and display.
Information, data, signal values, variable values, and parameters are stored in the main memory, registers, cache memory, and buffers during the CPU operations of extraction, search, reference, comparison, calculation, processing, editing, output, printing, and display. It is temporarily stored in a memory or the like.
The arrows in the flowcharts described in the first and second embodiments mainly indicate input / output of data and signals. The data and signal values are the memory of the RAM 914, the flexible disk of the FDD904, the compact disk of the CDD905, and the magnetic field. Recording is performed on a recording medium such as a magnetic disk of the disk device 920, other optical disks, mini disks, DVDs, and the like. Data and signals are transmitted online via a bus 912, signal lines, cables, or other transmission media.

  In addition, what is described as “˜unit” in the description of the first and second embodiments may be “˜circuit”, “˜device”, “˜device”, and “˜step”. , “˜procedure”, and “˜processing”. That is, what is described as “˜unit” may be realized by firmware stored in the ROM 913. Alternatively, it may be implemented only by software, or only by hardware such as elements, devices, substrates, and wirings, by a combination of software and hardware, or by a combination of firmware. Firmware and software are stored as programs in a recording medium such as a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, and a DVD. The program is read by the CPU 911 and executed by the CPU 911. That is, the program causes the computer to function as the “˜unit” in the first and second embodiments. Alternatively, the computer executes the procedure and method of “to unit” in the first and second embodiments.

  As described above, the mobile monitoring device 104 shown in the first and second embodiments includes a CPU as a processing device, a memory as a storage device, a magnetic disk, a keyboard as an input device, a mouse, a communication board, and a display device as an output device, A computer including a communication board or the like, and implements the functions indicated as “˜units” as described above using these processing devices, storage devices, input devices, and output devices.

1 is a diagram illustrating a configuration example of a monitoring system according to Embodiment 1. FIG. FIG. 3 shows an installation example of a monitoring camera or the like according to the first embodiment. FIG. 3 is a flowchart showing a video display flow at the time of intrusion detection according to the first embodiment. FIG. 3 is a flowchart showing a video display flow during tracking monitoring according to the first embodiment. The figure which shows the example of the database for travel time calculation which concerns on Embodiment 1. FIG. FIG. 5 is a diagram showing a tracking monitoring screen display example according to the first embodiment. The figure which shows the hardware structural example of the moving body monitoring apparatus which concerns on Embodiment 1,2.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 Network, 101 Camera, 102 Entrance / exit management apparatus, 103 sensor, 104 Moving body monitoring apparatus, 105 Video recording apparatus, 106 Monitoring display, 107 Camera changeover switch, 110 Camera moving body detection part in a camera image, 111 Inter-camera movement path Calculation unit, 112 Inter-camera moving body matching degree calculation unit, 113 ACS moving body detection unit, 114 Monitoring database, 115 Position management unit, 116 Video composition / display switching unit, 117 Sensor moving body detection unit, 118 Inter-camera path learning Unit, 119 behavior analysis intrusion determination unit, 120 camera image receiving unit, 121 monitoring target space, 122 monitoring target space, 123 monitoring target space.

Claims (9)

A captured video receiver that receives captured video from a plurality of cameras arranged in the monitoring target space;
Estimated shooting that selects one or more cameras that are estimated to be able to shoot the moving body when the moving body enters the monitoring target space and the moving body reaches the shootable range in the near future. Possible camera selection part,
An estimated arrival time calculation unit that calculates, for each estimated image-capable camera, a time estimated that the moving body reaches the image-capable range of the estimated image-capable camera;
A display unit configured to output to the display device a captured image of the estimated shootable camera received by the captured image reception unit and an estimated arrival time of each estimated shootable camera calculated by the estimated arrival time calculation unit; A monitoring device characterized by that. The monitoring device further includes:
An average moving time information storage unit storing average moving time information indicating the average moving time for each moving object attribute when moving between cameras;
The estimated arrival time calculation unit
2. The monitoring according to claim 1, wherein an estimated arrival time is calculated for each estimated image-capable camera using an average moving time for a moving object attribute that matches the attribute of the moving object that has entered the monitoring target space. apparatus. The estimated arrival time calculation unit
Compare the measured travel time required when moving from one camera to another camera that has moved into the monitored space, and the average travel time for each moving object attribute when moving between the cameras. A mobile object attribute whose time approximates to an actually measured travel time of a mobile object that has entered the monitoring target space is determined, and an estimated arrival time is estimated for each camera capable of being estimated using an average travel time for the determined mobile object attribute. The monitoring device according to claim 2, wherein the monitoring device is calculated. The average travel time information storage unit
For any combination of cameras in the plurality of cameras, storing average moving time information indicating an average moving time for each moving body attribute when moving between cameras that form a set,
The estimated photographing capable camera selection unit is
Select one or more cameras that are estimated to be able to shoot the moving body from the camera that is shooting the moving body, and that the moving body has reached the shootable range in the near future and is capable of shooting the moving body,
The estimated arrival time calculation unit
The estimated arrival time is calculated for each estimated photographable camera using the average movement time for the moving object attribute that matches the attribute of the moving object among the average moving time from the camera being photographed to each estimated photographable camera. The monitoring apparatus according to claim 2 or 3, wherein The average travel time information storage unit
The average movement time when a person who understands the movement path of the monitoring target space walks, the average movement time when a person who understands the movement path of the monitoring target space travels, and the monitoring target space 5. The average moving time information indicating at least one of the average moving times when a person who does not understand the moving route walks as the average moving time for each moving body attribute is stored. The monitoring apparatus in any one of. The photographed video receiver is
Receiving captured images from the first camera and the second camera arranged in the monitoring target space,
The monitoring device further includes:
Analyzing the captured video from the first camera to identify a specific moving body from the captured video from the first camera, and analyzing the captured video from the second camera to analyze the second camera A moving body identification unit for identifying the specific moving body from the captured video from
The time when the moving object identifying unit has identified that the specific moving object has departed from the shootable range of the first camera from the captured image from the first camera, and the moving object identifying unit The difference from the time when the specific moving body is identified as being in the shootable range of the second camera from the captured video from the second camera, from the first camera to the second camera. An average travel time calculation unit that calculates the average travel time;
The average time information storage unit
6. The monitoring apparatus according to claim 2, wherein average moving time information indicating the average moving time calculated by the average moving time calculating unit is stored. The monitoring device further includes:
The first entrance / exit management device or the first entry / exit management device arranged near the first camera and the time when the specific moving body passes the first sensor. Or acquiring information indicating the time when the specific moving body passes through the second sensor, the time when the specific moving body passes through the first entrance / exit management device or the first sensor, and the first An average moving time calculation unit that calculates a difference from the time when the specific moving body passes through a second entrance / exit management device or a second sensor as an average moving time from the first camera to the second camera Have
The average time information storage unit
6. The monitoring apparatus according to claim 2, wherein average moving time information indicating the average moving time calculated by the average moving time calculating unit is stored. A captured image receiving step in which the computer receives captured images from a plurality of cameras arranged in the monitored space;
When a moving object enters the monitoring target space, the computer selects one or more cameras that are estimated to be capable of shooting the moving object when the moving object reaches the imageable range recently. A step of selecting an estimated camera capable of shooting,
An estimated arrival time calculating step in which the computer calculates, for each estimated image-capable camera, a time estimated that the moving body reaches the image-capable range of the estimated image-capable camera;
A display step in which the computer outputs to the display device the captured video of the estimated shootable camera received in the captured video receiving step and the estimated arrival time for each estimated shootable camera calculated in the estimated arrival time calculating step. The monitoring method characterized by having. Captured video reception processing for receiving captured video from a plurality of cameras arranged in the monitoring target space;
Estimated shooting that selects one or more cameras that are estimated to be able to shoot the moving body when the moving body enters the monitoring target space and the moving body reaches the shootable range in the near future. Possible camera selection process,
Estimated arrival time calculation processing for calculating for each estimated photographable camera, as the estimated arrival time, the estimated time when the mobile body reaches the photographing possible range of the estimated photographing camera;
A display process for outputting, to a display device, a captured image of the estimated photographable camera received by the captured image reception process and an estimated arrival time for each estimated photographable camera calculated by the estimated arrival time calculation process. A program characterized by being executed.

Images