JP5386744B2 - Monitoring system - Google Patents

Description

  The present invention relates to a monitoring system that monitors an area to be monitored, and more particularly to a monitoring system that automatically detects a person who falls from a platform to a track at a station.

For example, when a person falls from the platform of a station to the track, the presence of the fallen person is promptly transmitted to the driver of a train or other vehicle entering the platform, and the vehicle is stopped to ensure safety. There is a need. The detection and transmission of a fallen person by visual inspection is a heavy burden on the observer, and it takes time for the observer to detect the fallen person and transmit it using a communication device. In fighting situations it can lead to danger.
Therefore, attempts have been made to automatically detect a fallen person by using an external sensor such as a mat sensor or a wire sensor. However, since a train passes on the track, these sensors can be installed in places where they can be installed. There is a limit, and there are problems such as oversight when a fallen person falls outside the sensor installation location.

  For this reason, a method of applying a technique for detecting an object entering a specified area by image processing to a fallen person from a station platform can be considered. According to the image processing, it is possible to reduce oversight caused by the installation position like a mat sensor or a wire sensor. In such object detection technology based on image processing, a so-called difference method for detecting an object based on a difference between an input image and a background image is widely used.

  Here, the difference method is a difference image obtained by calculating a difference for each pixel between an input image obtained from an imaging device or the like and a background image on which an object to be detected that is prepared is not reflected. In this method, each pixel of an image is compared with a threshold value (for example, 20) prepared in advance, and an object to be detected is found from the position of a pixel having a difference value equal to or greater than the threshold value. However, in the fallen person detection based on the difference method, there is a possibility that disturbance noises other than the fallen person may be erroneously detected, such as a shadow of a person existing on the platform, a train approaching the platform, or a headlight thereof.

JP 2002-157599 A JP 2007-233919 A Japanese Patent No. 4087045 JP 2006-82618 A Japanese Patent No. 3607653 JP 2008-176768 A

As described above, in a monitoring system for monitoring an object that enters an area to be monitored using an imaging device, when detecting a fallen person from a home to a track at a station or the like, from the home to the track by monitoring based on image processing. In the monitoring of the fallen person, there is a problem that a shadow of a person existing on the platform or a disturbance such as a train or a headlight entering the home may be erroneously detected as a fallen person.
The present invention has been made in view of such a conventional situation, and performs image processing of a video (image) of a monitoring camera that images a home and a track, distinguishes disturbances from a home monitoring from a fallen person, and falls only. It is an object of the present invention to provide a monitoring system that can accurately detect and issue a report.

In order to achieve the above object, in the monitoring system according to the present invention, the presence / absence of a fallen person from the station platform is determined by the following configuration.
In other words, the imaging means captures an image of a monitoring target area related to the station platform and the track. A boundary line setting means sets a boundary line between the home and the track. The feature condition setting means sets a condition relating to a feature for determining whether or not the object image is a fallen person image. The object detection unit detects an image of the object included in the image captured by the imaging unit. The feature detection means detects the feature of the image of the object detected by the object detection means. The feature condition determining unit determines whether or not the feature detected by the feature detecting unit satisfies a condition set by the feature condition setting unit. The position change detection means detects a change in the position of the image of the object detected by the object detection means. The boundary line determination unit determines whether or not the image of the object detected by the object detection unit crosses the boundary line set by the boundary line setting unit based on the detection result by the position change detection unit. The fallen person determination means determines that the feature detected by the feature detection means by the feature condition determination means satisfies the condition set by the feature condition setting means, and the object detection by the boundary line determination means When it is determined that the image of the object detected by the means crosses the boundary set by the boundary setting means, the image of the object detected by the object detection means is an image of a fallen person By determining, the presence or absence of a fallen person is determined.

  Accordingly, it is possible to improve the accuracy of determining whether or not there is a fallen person from the platform of the station. It is possible to detect only the person accurately and issue a report.

Here, various areas for monitoring the presence or absence of a fallen person may be used as the area to be monitored regarding the platform and the railway line of the station. For example, a part or the whole of the platform and a railway line adjacent to the platform. An area including (a part of a track extending over a plurality of stations) can be used.
In addition, various boundary lines may be used as the boundary line between the home and the track, and various lines such as a straight line and a curve may be used as the line.

Various features may be used for determining whether the image of the object is a fallen person's image. For example, a size such as a width or a height, or a speed may be used. Can do.
In addition, as conditions regarding such features, for example, “more than a predetermined value (or larger than a predetermined value)”, “below a predetermined value (or smaller than a predetermined value)”, “within a predetermined range (or A condition such as “outside of a predetermined range” can be used.

Various methods may be used as a method for setting information such as boundary lines and conditions. For example, a method for setting information according to a human operation, or an initial setting in a system (for example, a certain device) in advance. Can be used. The set information is stored in a memory of a system (for example, a certain device), for example.
The image of the object included in the captured image is based on, for example, a comparison (difference or the like) with a background image when the object does not exist, or a result of searching a template of the object in the image. Can be detected.

Further, as a change in the position of the image of the object, for example, a trajectory along which the object has moved can be detected.
In addition, regarding the determination of whether or not the object image has crossed the boundary line, it is possible to use a mode in which the direction across the boundary line is set. In this case, the object image crosses the boundary line in the set direction. Only when the object image crosses the boundary line, and even if the object image crosses the boundary line in a different direction, it is not determined that the object image crosses the boundary line.

The monitoring system according to the present invention has the following configuration as a configuration example.
That is, the vehicle status determination means determines the status of the presence of the vehicle in the monitored area. When the determination stop means determines that the situation exists from the situation in which no vehicle exists in the monitored area based on the determination result by the vehicle situation determination means (for example, when the vehicle has entered) ) Stops the determination of the presence or absence of a fallen person. The determination restarting means does not exist from the situation in which the vehicle exists in the monitored area based on the determination result by the vehicle status determination means in the state where the determination of the presence or absence of the fallen person is stopped by the determination stop means. When it is determined that the situation has occurred (for example, when the vehicle has left), the determination of the presence or absence of a fallen person is resumed.

  Therefore, for example, when the vehicle has entered and stayed in the monitored area, the determination of the presence or absence of a fallen person is stopped, and then the determination of the presence or absence of a fallen person is resumed when the vehicle leaves the monitored area. Thus, it is possible to prevent the vehicle from being erroneously detected as a fallen person.

Here, various situations may be determined as the situation of the presence of the vehicle in the monitoring target area. For example, a configuration in which one or more situations such as entry, stay, and exit of the vehicle are determined can be used. .
In addition, various methods may be used as a method for determining the vehicle presence state in the monitoring target region, for example, a predetermined signal including information indicating the vehicle presence state in the monitoring target region. A method of receiving (for example, an external signal) and determining based on the content of the signal, or an image of an area to be monitored (for example, a real-time image) and a background image when no vehicle is present A method of determining based on the difference or the like, or a method of using both of these methods in combination can be used.

In addition, as an aspect of stopping the determination of the presence or absence of a fallen person, for example, an aspect of stopping all the processes related to the determination of the presence or absence of a fallen person may be used, or For example, a mode in which some processes such as detection of the feature of the image of the object and detection of a change in position are performed without stopping may be used.
Moreover, as an aspect which restarts determination of the presence or absence of a fallen person, for example, all the stopped processes are restarted.

  As described above, according to the monitoring system of the present invention, it is possible to improve the accuracy of the determination of the presence or absence of a fallen person from the station platform.

It is a figure which shows the structural example of the monitoring system which concerns on one Example of this invention. It is a figure which shows an example of the input image imaged by the imaging device. It is a figure which shows an example of the image which performed the setting for detecting approach to a track from a home. It is a figure which shows an example of the output image displayed on a display apparatus when a fallen person is detected. It is a figure which shows an example of the input image when the vehicle has approached the monitoring area | region. It is a figure which shows an example of the setting for judging the presence or absence of the approach of a vehicle. It is a figure which shows an example of the input image at the time of vehicle approach. It is a figure which shows an example of the input image when a bird enters a rectangle.

Embodiments according to the present invention will be described with reference to the drawings.
In the present embodiment, various image processing such as threshold value and line setting processing for an image, and detection (detection) processing of a predetermined target in the image, its rectangular area, position change (trajectory), etc. Various methods may be used, and various known techniques may be used. For example, the difference method or the like is a known technique (see, for example, Patent Document 6).

FIG. 1 shows a configuration example of a monitoring system according to an embodiment of the present invention.
The monitoring system of this example includes an imaging device 1, an object detection device 2, an instruction device 3, an alarm output device 4, and a display device 5.
The object detection apparatus 2 includes a video input circuit 11, an image processor 12, a program memory 13, a work memory 14, an external I / F (interface) circuit 15, a video output circuit 16, a data bus ( (Communication path) 17.

The imaging device 1 is configured using, for example, a camera and images a region to be monitored. The captured input image is recorded in the work memory 14 via the video input circuit 11 in the object detection device 2.
In the object detection device 2, for example, an input image captured by the imaging device 1 is processed to extract an object in the monitoring area, and it is determined whether or not the extracted object is a target to be detected such as a fallen person. Detect an object in the monitoring area.
The setting device (indicating device 3 in this example) adjusts parameters used in image processing or the like in the object detection device 2.
The alarm output device 4 outputs an alarm based on the detection result in the object detection device 2 and the like.
The display device 5 displays the output image on the screen based on the detection result in the object detection device 2 and the like.

  In the object detection device 2, the image processor 12 processes the input image recorded in the work memory 14 in accordance with the program recorded in the program memory 13, and displays the processing result via the video output circuit 16. 5 is displayed. Further, the image processor 12 outputs an alarm signal via the external I / F circuit 15 when it is determined that the alarm should be issued as a result of the image processing, whereby the alarm output device Alarm is issued according to 4.

In addition, the image processor 12 sets the parameters of the above-described program based on an instruction from an operator (person) using the pointing device 3 such as a mouse or a keyboard, which is input via the external I / F circuit 15. The input image is processed while changing or correcting.
The video input circuit 11, the image processor 12, the program memory 13, the work memory 14, the external I / F circuit 15, and the video output circuit 16 are connected to a data bus 17.

FIG. 2 shows an example of an input image imaged by the imaging device 1.
In the monitoring system of this example, the object is to automatically detect a person entering the vicinity of the track 102 from the home 101 as a fallen person and issue an alarm.

FIG. 3 shows an example of an image that has been set to detect entry from the platform 101 to the track 102.
Specifically, the direction indicated by the detection line 111 and the arrow 112 is set. In this example, since the fallen person moves from the home 101 to the track 102, only the object that exceeds the detection line 111 in the direction indicated by the arrow 112 is detected. This can reduce false detections such as shadows and changes in sunshine.

Furthermore, in this example, in order to increase the identification accuracy of whether or not the detected object is a fallen person to be notified, the detected object is detected only when the size and speed of the detected object are within a certain range. Judged as the target of the alarm.
In this example, as an example, the size and speed that can be assumed when one person falls freely from a 2 m height is used as the condition of the fallen person. As a specific example, the detected object has a width of 0.5 m to 2 m. Only when the height is within a range of 0.0 m, a height of 0.5 m to 2.0 m, and a speed of 3.0 m / s to 7.0 m / s, the object is determined to be a fallen person.
Here, the conditions such as the detection line 111, the direction (in this example, the arrow 112), and the size and speed of the fallen person are arbitrarily set by the operator using the pointing device 3 in the object detection device 2. Can do.

FIG. 4 shows an example of an output image displayed on the display device 5 when a fallen person is detected.
For the fallen person 121 found by the detection, the object detection device 2 adds a rectangle 122 indicating the position of the faller 121 and a trajectory 123 indicating the movement history to the input image and outputs them. In this example, the rectangle 122 and the trajectory 123 are displayed only for objects that exceed the detection line 111 in the direction of the arrow 112, particularly those whose size, speed, etc. meet the above-mentioned conditions of the fallen person. Is done. At the same time that the fallen person 121 is detected, an alarm signal is transmitted to the alarm output device 4 to generate an alarm.

When an alarm is issued, an output image may be transmitted remotely to a display device such as a driver via a wireless network, etc., so that the driver can easily visually confirm the state of the site.
In addition, if the approach of a fallen person does not depend on one direction from the home, a method of installing a plurality of detection lines or a candidate of a fallen person who should detect an object that has entered a railroad crossing regardless of the approaching direction A method of setting a detection area for the detection line instead of the detection line may be used.

In addition, since a vehicle such as a train enters the monitoring area in monitoring the fallen person from the home 101, there is a possibility that the vehicle is erroneously detected as a fallen person. In this example, the train stays in the monitoring area. During this time, the object detection process is stopped.
FIG. 5 shows an example of an input image when the vehicle 131 enters the monitoring area.
Furthermore, since the vehicle 131 may be reflected in the background image in the background subtraction method and an abnormal background image may be generated by stopping the vehicle 131 for a certain time, in this example, when the vehicle 131 enters and exits, the background Do not update the image.

  Here, various methods may be used as a method for judging the entry and exit of the vehicle. For example, an external signal having information on entry and exit of the vehicle such as a crossing signal is sent from the external I / F circuit 15. When the signal that informs the entry of the vehicle is inputted, the object detection process and the update of the background image are stopped, and further, the signal that informs the exit of the vehicle is inputted. Then, the object detection process and the update of the background image are resumed. There may be a time difference between the input of such a signal and the actual entry of the vehicle into the monitoring area of the input image depending on the signal specifications and the camera installation location. Can be predicted, the object detection process and background image update stop and restart should be shifted by the predicted time.

  Moreover, when the signal which conveys approach of a vehicle cannot be used, or such a signal can be used, but the time difference of the input of a signal and the actual vehicle approach area is estimated. If this is not possible, for example, template matching can be performed between the input image and the background image in a specific area, and the method can be replaced by determining whether the vehicle has entered based on the magnitude of the correlation value. It is.

This will be described with reference to FIGS.
FIG. 6 shows an example of a setting for determining whether or not a vehicle has entered.
In this example, one or more regions such as a rectangle 141 and a rectangle 142 are drawn by the setting device (in this example, the indication device 3), and the input image and the background image are represented by the rectangles 141 and 142, respectively. Are measured by template matching such as SSD (Sum of Squared Difference) method.

  As a result of this measurement, for example, it is determined that the vehicle has entered when the correlation value falls below a threshold value in one of the rectangular 141 and rectangular 142 regions. In this example, the threshold value of the correlation value for determining that the vehicle has entered is 0.5. In the example of FIG. 6, since no vehicle or the like has entered the range of the rectangles 141 and 142, the input image is similar to the background image (in this example, an image when no vehicle is present). The correlation value between the input image and the background image indicates a high value such as 0.95 for the rectangle 141, 0.90 for the rectangle 142, etc., and the object detection process and the background image update are continued.

FIG. 7 shows an example of an input image at the time of entering the vehicle.
When a vehicle or the like enters the range of the rectangles 141 and 142, the pattern of the input image and the background image are greatly different. For example, the correlation value between the input image and the background image in the rectangle 141 is a low value such as 0.3. Show. At this time, since the correlation value is below the threshold value of 0.5, it is determined that the vehicle has entered the monitoring area, and the object detection process and the background image update are stopped.

In addition, in the determination of vehicle entry by template matching or the like, for example, when an object other than a vehicle such as a small animal enters the rectangles 141 and 142, the correlation value decreases, so that erroneous determination of vehicle entry may occur. There is sex.
FIG. 8 shows an example of an input image when a bird 151 enters a rectangle 142.
At this time, if the correlation value becomes 0.4 due to the entry of the bird 151 in the rectangle 142, the correlation value falls below the threshold value, so it is erroneously determined that the vehicle has entered and the vehicle has not entered. Therefore, the object detection process and the background update process are erroneously stopped.

Therefore, as an effective method for reducing the frequency of erroneous determination of vehicle entry, in this example, a plurality of rectangles are linked and regarded as vehicle entry only when the correlation value fluctuates in a specific order, etc. The method is used.
For example, when the vehicle travels from the back to the front of the screen, the rectangle 141 reaches the vehicle earlier than the rectangle 142, so the correlation value also falls below the threshold value in the order of the rectangle 141 and the rectangle 142. Therefore, the determination of vehicle entry also operates only when the correlation value falls below the threshold value in this order (considered as vehicle entry). For example, when the rectangle 142 alone falls below the threshold value The vehicle entry determination condition is changed by combining a plurality of rectangles 141 and 142 such that the vehicle is not regarded as a vehicle entry. As a result, even when the bird 151 enters the rectangle 142, since the fluctuation is small in the rectangle 141 and the correlation value does not fall below the threshold value, it is realized that the vehicle is not regarded as entering the vehicle.

  In addition, immediately after an object predicted to be a fallen person passes through the detection line 111, when the correlation value of the rectangle 141 or the rectangle 142 falls below the threshold value, it is confirmed that the faller has entered the rectangles 141 and 142. For example, the object detection process is not stopped because the possibility is high. In addition, as the vehicle entry determination, by performing in combination with external information, such as performing a determination that the vehicle is approaching by an external signal or only a few minutes before and after that time, It is possible to reduce malfunction of vehicle entry determination.

As described above, in this example, the input image obtained by imaging the monitoring area by the imaging device 1 having the imaging lens and the imaging element is processed, and the object in the monitoring area is extracted by extracting the object from the input image. In the monitoring system to be detected, the following processing is performed.
That is, a virtual home / line boundary line (in this example, the detection line 111) is set on the video obtained by imaging the video of the home 101 and the line 102 at the station by the imaging device 1 by the detection line setting function. By the object tracking function, a change in the position of the detected object (in this example, a locus 123) is obtained. The object feature extraction function extracts at least one feature of the object such as the size and speed of the detected object. The line intersection determination function determines whether or not the object has crossed the home / line boundary set by the detection line setting function based on the change in the position of the object obtained by the object tracking function. The fall determination function determines the presence or absence of a fallen person based on the object features extracted by the object feature extraction function and the determination result by the line intersection determination function (determination result whether or not the object crosses the boundary line). .

  With these functions, for example, an object having a predetermined range of size (or a predetermined range of speed, etc.) moved to the track side across the boundary between the home 101 and the track 102 is judged as a fallen person. Automatically detect falls in the area. Thereby, for example, it is possible to identify noise and a fallen person and detect only the fallen person in the monitoring area (or increase the certainty thereof).

In the monitoring system of this example, the following processing is performed.
That is, the external signal input function inputs information about the presence or absence of the vehicle entering the monitoring area using an external signal. The fall determination stop function stops the above-described determination of the presence or absence of a fallen person when the vehicle enters or stays in the monitoring area. The fall determination restart function restarts the above-described determination of the presence or absence of a fallen person after leaving the vehicle.

Further, in the monitoring system of this example, the following processing is performed.
That is, the correlation value between the input image and the background image is calculated by the correlation calculation function when the vehicle enters the monitoring area (for example, when notified by an external signal). When the correlation value is lower than a predetermined value by the vehicle determination function (or when the correlation value is equal to or lower than the predetermined value), it is determined that the vehicle has entered and stayed in the monitoring area. And when it determines with the vehicle determination function not having approached the vehicle into a monitoring area | region, the above-mentioned fall determination function determines the presence or absence of a fallen person.

  Therefore, in the monitoring system of this example, for example, a person who falls from the home to the track can be accurately detected with few false alarms. This makes it possible for surveillance personnel and vehicle drivers to immediately check the real-time video of the monitoring area where the vehicle has fallen, reducing the number of false alarms and improving system reliability, and reducing labor. It is possible to monitor the fall at, and to perform quick danger avoidance behavior.

In the monitoring system of this example, an imaging unit is configured by the function of the imaging device 1, and a boundary line (in this example, a detection line) setting unit and a characteristic condition (characteristics) are configured by the functions of the pointing device 3 and the object detection device 2. In this example, width, height, and speed conditions) setting means are configured, and the object detection means for the input image and the object by the function of the image processor 12 of the object detection device 2 processing the input image. Image feature detection means, feature condition determination means for determining whether the image feature of the object satisfies the condition, position change detection means for the object image, and the object image crosses the boundary line A boundary line determination means for determining whether or not a person has fallen (in this example, a crossing direction is also determined) and a fallen person determination means for determining the presence or absence of a fallen person based on the determination results of features and boundary lines. .
Further, in the monitoring system of this example, the vehicle state determination means is configured by the function of the object detection device 2 determining the vehicle presence state in the monitoring area by an external signal or image processing, and object detection is performed according to the determination result. A determination stop unit and a determination restart unit are configured by the function of the device 2 that stops or restarts the determination of the presence or absence of a fallen person.

Here, the configuration of the system and apparatus according to the present invention is not necessarily limited to the configuration described above, and various configurations may be used. The present invention can also be provided as, for example, a method or method for executing the processing according to the present invention, a program for realizing such a method or method, or a recording medium for recording the program. It is also possible to provide various systems and devices.
The application field of the present invention is not necessarily limited to the above-described fields, and the present invention can be applied to various fields.
In addition, as various processes performed in the system and apparatus according to the present invention, for example, the processor executes a control program stored in a ROM (Read Only Memory) in hardware resources including a processor and a memory. A controlled configuration may be used, and for example, each functional unit for executing the processing may be configured as an independent hardware circuit.
The present invention can also be understood as a computer-readable recording medium such as a floppy (registered trademark) disk or a CD (Compact Disc) -ROM storing the control program, and the program (itself). The processing according to the present invention can be performed by inputting the program from the recording medium to the computer and causing the processor to execute the program.

1 .... Imaging device 2 .... Object detection device 3 .... Indicating device 4 .... Alarm output device 5 .... Display device 11 .... Video input circuit 12 .... Image processor 13 ... Program Memory, 14.Work memory, 15.External I / F circuit, 16.Video output device, 17.Data bus (communication path),
101 ························································································· •••••••••••••••••••••••••••••••••••••••••• Bird ,

Claims (2)

A monitoring system that determines the presence or absence of a fallen person from a station platform,
An imaging means for capturing an image of an area to be monitored related to the station platform and track;
Boundary line setting means for setting a boundary line between the home and the track;
This is a condition related to the characteristics for determining whether the image of the object is an image of a fallen person, and the size or speed of the object that is assumed when the object that has fallen freely from a predetermined height is a human being Characteristic condition setting means for setting at least one condition ;
Object detection means for detecting an image of an object included in the image captured by the imaging means;
Feature detection means for detecting features of an image of the object detected by the object detection means;
Feature condition determination means for determining whether the feature detected by the feature detection means satisfies a condition set by the feature condition setting means;
Position change detection means for detecting a change in the position of the image of the object detected by the object detection means;
Boundary line determination means for determining whether the image of the object detected by the object detection means crosses the boundary line set by the boundary line setting means based on the detection result by the position change detection means;
The feature detected by the feature detection unit by the feature condition determination unit is determined to satisfy the condition set by the feature condition setting unit, and the object detected by the object detection unit by the boundary line determination unit Is determined to be a fallen person's image by determining that the image of the object detected by the object detection means is an image of a fallen person. A fallen person judging means for judging the presence or absence of a person,
For the background image prepared for the monitoring target area and the image captured by the imaging means, template matching is performed for each of the plurality of partial areas in the image, and a correlation value between the images is calculated for each partial area. Based on the correlation value for each area, the situation of the presence of the vehicle in the area to be monitored is determined. At this time, when partial areas having correlation values below the threshold value occur in a specific time-series order, the monitoring object Vehicle status determination means for determining that a vehicle exists in the area of
Based on the determination result by the vehicle state determination means, when it is determined that the vehicle is not present in the monitored area, the determination stop is made to stop the determination of the presence or absence of a fallen person. Means,
In the state where the determination of whether or not there is a fallen person has been stopped by the determination stop means, the situation has changed from the situation in which a vehicle is present in the monitored area based on the determination result by the vehicle situation determination means. Is determined, a determination restarting means for restarting the determination of the presence or absence of a fallen person,
A monitoring system characterized by comprising: A monitoring method implemented by a monitoring system for determining the presence or absence of a fallen person from a station platform,
  The monitoring system includes an imaging unit that captures an image of an area to be monitored related to a station platform and a track,
  Detecting an image of an object included in an image captured by the imaging means;
  Detecting changes in the image characteristics of the detected object and the position of the image of the object;
  This is a condition related to the characteristics for determining whether the image of the object is an image of a fallen person, and the size or speed of the object that is assumed when the object that has fallen freely from a predetermined height is a human being Determining whether or not the image feature of the detected object satisfies a condition set for at least one of the conditions;
  Based on the change in the position of the detected object image, it is determined whether the detected object image has crossed the boundary set for the home and the track;
  When it is determined that the image feature of the detected object satisfies the set condition, and it is determined that the image of the detected object crosses the set boundary line, By determining that the image of the detected object is an image of a fallen person,
  For the background image prepared for the monitoring target area and the image captured by the imaging means, template matching is performed for each of the plurality of partial areas in the image, and a correlation value between the images is calculated for each partial area. Based on the correlation value for each area, the situation of the presence of the vehicle in the area to be monitored is determined. At this time, when partial areas having correlation values below the threshold value occur in a specific time-series order, the monitoring object It is determined that there is a vehicle in the area of
  Based on the determined situation of the presence of the vehicle, when it is determined that the situation exists from the situation in which no vehicle exists in the monitored area, the determination of the presence or absence of a fallen person is stopped, In the state where the determination of the presence or absence of a fallen person has been stopped, if it is determined that the situation does not exist from the situation where a vehicle exists in the monitored area, the determination of the presence or absence of a fallen person is resumed.
  A monitoring method characterized by that.