JP6810977B2 - Monitoring system, monitoring method, monitoring program and its recording medium - Google Patents

Description

The present invention relates to a monitoring system, a monitoring method, a monitoring program, and a recording medium thereof.

Conventionally, in hospitals and nursing homes, there are devices and methods for monitoring the behavior of inpatients and facility residents and predicting the behavior of inpatients and facility residents based on the monitoring results.

For example, in Japanese Patent Application Laid-Open No. 2014-166197 (Patent Document 1), a camera photographs a bed and the vicinity of the bed, and the positional relationship between an object such as a bed and a moving body region in which movement is occurring is determined. An information processing device including a behavior estimation unit that estimates the behavior of the person to be watched over the object is described. Then, when the presumed behavior of the watching target person is an action that indicates a sign that the watching target person is in danger, it is described that the person watching the watching target person is notified to notify the sign.

Further, in Japanese Patent Application Laid-Open No. 2014-174627 (Patent Document 2), a moving object is detected and detected based on an optical flow in the moving object by shooting a moving image of the behavior of a person to be watched on or near the bed. Described is an information processing apparatus including a behavior estimation unit that estimates the behavior of a person to be watched on or near the bed based on the position and the moving direction of the moving object.

Japanese Patent Application Laid-Open No. 2014-182409 (Patent Document 3) describes the environment in which the monitored person to be photographed is placed according to the installation position of the photographing unit in order to analyze the image according to the installation position of the photographing unit. A monitoring device that stores the monitoring environment data specified in advance is described.

In Japanese Patent Application Laid-Open No. 2014-236896 (Patent Document 4), the depth is obtained for each pixel of a photographed image of a person to be watched and an object such as a bed which is a reference for the behavior of the person to be watched. An information processing device that estimates the positional relationship between the person to be watched and the subject including the object in the real space and estimates the behavior of the person to be watched is described.

Japanese Unexamined Patent Publication No. 2014-166197 Japanese Unexamined Patent Publication No. 2014-174627 Japanese Unexamined Patent Publication No. 2014-182409 Japanese Unexamined Patent Publication No. 2014-236896

However, in the information processing apparatus of Japanese Patent Application Laid-Open No. 2014-166197 (Patent Document 1), only the behavior of the watching target person with respect to one specific object such as a bed can be estimated. In addition, the behavior of the person being watched over is estimated uniformly for any person being watched over based only on the positional relationship between the person being watched over and the moving object, so it is correct depending on the characteristics of the person being watched over. It may not be estimated correctly, or it may not be estimated correctly if the characteristics or environment of the person being watched over changes.

Further, even in the information processing apparatus of Japanese Patent Application Laid-Open No. 2014-174627 (Patent Document 2), how is it based on the position and the moving direction of the moving object detected in one limited range such as on or near the bed? Since the behavior of the watching target person is estimated uniformly even for the watching target person, it is correctly estimated when it is not estimated correctly depending on the characteristics of the watching target person, or when the characteristics or environment of the watching target person changes. It may disappear.

In the monitoring device of Japanese Patent Application Laid-Open No. 2014-182409 (Patent Document 3), although the position of the photographing unit can be changed, the standard of the object to be monitored is like a bed in the environment stored in the monitoring environment data. It belongs to one specific place, and only the operation of the monitored person in the specific place can be monitored. In addition, this monitoring device does not estimate the behavior of the monitored person or predict the danger.

In the information processing device of Japanese Patent Application Laid-Open No. 2014-236896 (Patent Document 4), the behavior of the watching target person in the real space instead of the two-dimensional image is estimated based on the two-dimensional photographed image. It's a specific place like a bed. Similarly, even with this information processing device, it may not be estimated correctly depending on the characteristics of the person being watched over, or it may not be estimated correctly when the characteristics or environment of the person being watched over are changed.

When the behavior estimated for the watch target person is an action that indicates a sign that the watch target person is in danger, the behavior of the watch target person is performed in the monitoring system that is configured to notify the watcher of the watch target person. If it is not detected correctly and estimated correctly, the risk that may reach the person being watched over cannot be predicted correctly.

Therefore, an object of the present invention is to provide a monitoring system, a monitoring method, a monitoring program, and a recording medium thereof, which can improve the accuracy of prediction of danger to a monitored person.

In the monitoring system according to the present invention, as an element for identifying the event of the monitored person, the area where the monitored person exists, the detection unit that detects the status of the monitored person, and the detection unit are detected. The judgment unit is equipped with a judgment unit that determines the event of the monitored person based on the event element, and a storage unit that stores at least the combination of the area and status in which the monitored person exists and the event of the monitored person. , At least the area detected by the detection unit and the change in status are collated with the combination stored in the storage unit to determine the event of the monitored person.

Even if the status of the monitored person is the same, whether or not the monitored person is at risk may differ depending on the area where the status is taken. Therefore, by determining the event of the monitored person based on not only the status of the monitored person but also the status and the area, it is possible to improve the accuracy of predicting the danger to the monitored person.

In the monitoring system according to the present invention, the detection unit is further configured to detect the presence or absence of breathing of the monitored person, and the combination stored in the storage unit is further associated with the presence or absence of breathing of the monitored person. It is preferable that the determination unit is further configured to determine the event of the monitoring target person by collating with the combination associated with the presence or absence of breathing of the monitoring target person detected by the detection unit.

By doing so, it is possible to prevent the monitored person from losing sight of the dangerous state even when there is no major movement, such as a monitored person resting on the bed.

In the monitoring system according to the present invention, the detection unit is configured to further detect the time, the combination stored in the storage unit is further associated with the time, and the determination unit is further detected by the detection unit. It is preferable that the time is matched with the combination to determine the event of the monitored person.

For example, in the area on the bed, if the status of the monitored person changes from "on the bed" to "get up", there is no problem during the daytime, but at midnight, to deal with the toilet or wander around. May need to be dealt with. Therefore, by determining the event based on the time in this way, the accuracy of predicting the danger to the monitored person can be further improved.

In the monitoring system according to the present invention, the area detected by the detection unit preferably includes a bed, a living room, a toilet, an entrance / exit, and an outside of the living room.

In the monitoring system according to the present invention, the status of the monitored person preferably includes lying down, standing, getting up, boundary position, getting out of bed, entering and leaving the room, falling, crouching, and lying down.

The monitoring system according to the present invention preferably includes a notification unit that notifies the observer of an event determined by the determination unit.

The monitoring system according to the present invention includes an input unit in which the monitor inputs the correctness of the event notified by the notification unit to the judgment unit, and the judgment of the event of the monitored person by the judgment unit is input by the input unit. It is preferable that it is configured to be updated based on correctness.

By doing so, the observer can give feedback to the judgment by the judgment unit, so that the behavior of the monitored person can be estimated according to the characteristics of the watched person, the characteristics of the watched person, and changes in the environment. Therefore, the accuracy of predicting the danger to the monitored person can be improved.

In the monitoring method according to the present invention, the information processing apparatus detects the area where the monitored person exists and the status of the monitored person as elements for identifying the event of the monitored person. A step to determine the event of the monitored person based on the elements of the event, a step to memorize at least the combination of the area where the monitored person exists and the status with the event of the monitored person, and at least the detected area. Includes a step to determine the monitored person's event by matching the and status with the memorized combination.

The monitoring program according to the present invention has, as elements for identifying the event of the monitored person, the step of detecting the area where the monitored person exists, the status of the monitored person, and the element of the detected event. The step of determining the event of the monitored person based on it, the step of memorizing at least the combination of the area and status where the monitored person exists and the event of the monitored person are memorized, and at least the detected area and status are memorized. It is a monitoring program for causing an information processing apparatus to execute a step of determining an event of a monitored person by collating with the combined combination.

The recording medium according to the present invention is a recording medium on which the above program is recorded.

By doing so, it is possible to provide a monitoring system, a monitoring method, a monitoring program, and a recording medium thereof, which can improve the accuracy of predicting the danger of the monitored person.

It is a block diagram which shows the control-related configuration of the monitoring device of this invention and the device connected to the monitoring device. It is a figure which shows typically an example of the map of the room where the monitoring device of this invention is installed. As an example of a predetermined pattern, a diagram (A) schematically showing a time change of displacement in the first region when the monitored person is resting in the bed area, and a time of displacement in the second region are shown. It is a figure (B) which shows the change schematically. It is a figure which shows an example of the matrix which associated the status, area, and a biological detection result with an event.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

As shown in FIG. 1, in this embodiment, the monitoring system 1 is mainly composed of a monitoring device 100, a server 200, and a monitor terminal 300.

The monitoring device 100 is installed, for example, in the living room of the person to be monitored. The monitoring system 1 may include a plurality of monitoring devices 100, for example, the number of living rooms to be monitored. The monitoring device 100 is, for example, an information processing device including an external connection unit 110, a control unit 120, a storage unit 130, and an infrared sensor 140 that connect the server 200 and the monitoring device 100. The control unit 120 of the monitoring device 100 is composed of a CPU, RAM, ROM, and the like. The monitoring device 100 may include a plurality of processors.

In this embodiment, the storage unit 130 stores the program 131, the matrix 132, the map 133, and the pattern 134. The program 131 is a monitoring program 131 for causing the monitoring device 100 to execute the monitoring method. The matrix 132 is a combination of the status, area, biometric detection result, and time, which will be described in detail later, in association with the event of the monitored person. In the map 133, in this embodiment, each area of the bed, the living room, the toilet, and the doorway is designated as the area in the image shooting target area in the room of the monitored person. Pattern 134 is a temporal change in the tremor of the body at rest of the monitored subject. Since the matrix 132, the map 133, and the pattern 134 are updated, the storage unit 130 may be configured to store only the latest matrix 132, the map 133, and the pattern 134.

The control unit 120 executes the program 131 stored in the storage unit 130 to control each component. When the control unit 120 executes the monitoring method program 131, the monitoring device 100 includes the human body detection unit 121 including the area detection unit 121a and the status detection unit 121b, the first measurement unit 122a, and the second measurement unit 122b. It functions as a monitoring device 100 including a biological detection unit 122, a clock 123, and a determination unit 124 including the above. The function of each part will be described in detail later. The human body detection unit 121, the biological detection unit 122, and the clock 123 of the control unit 120 are collectively referred to as a detection unit.

The server 200 includes, for example, an external connection unit 210, a clock 220, a drive 230, and a storage unit 240. The external connection unit 210 is for connecting the server 200 and the monitoring device 100, and the server 200 and the monitor terminal 300. In this embodiment, the clock 220 of the server 200 is configured to be able to acquire accurate time information. The drive 230 is for reading the recording medium on which the monitoring method program 131 is recorded when the monitoring method program 131 is not stored in the storage unit 130 of the monitoring device 100. Therefore, the program 131 may be recorded on the recording medium read by the drive 230 instead of the storage unit 130 of the monitoring device 100. As a recording medium for recording the program 131, a known medium such as a CD, a DVD, or a semiconductor memory is used. The storage unit 240 of the server 200 may store, for example, not only the latest program 131, matrix 132, map 133, and pattern 134, but also the update history of the matrix 132, map 133, and pattern 134.

The observer terminal 300 is a terminal for the observer to operate at hand. The monitoring system 1 may include a plurality of observer terminals 300, for example, the number of observers. The observer terminal 300 includes, for example, an external connection unit 310, a clock 320, an input unit 300, a notification unit 340, and a display unit 350. The external connection unit 310 is for connecting the monitor terminal 300 and the server 200. The input unit 330 is used by the observer to input information to the control unit 120 of the monitoring device 100, and is configured as an input device such as a mouse, a keyboard, or a touch panel, for example. The notification unit 340 is for notifying the observer, for example, may be configured as a display for displaying characters or images, or may be configured to notify a specific event by turning on or off the light. It may be configured to notify a specific event by generating or stopping a sound, or it may be configured in other ways. Further, the notification unit 340 may be configured to notify only when the event is a sign of danger. The notification unit 340 may be configured to notify the event not only to the observer but also to the monitored person.

Omission, replacement, or addition of each component of the monitoring system 1 is optional. Further, the monitoring device 100, the server 200, and the monitoring terminal 300 of the monitoring system 1 may each be configured by a plurality of device groups.

Next, the operation of the monitoring system 1 configured in this way will be described. The monitoring system 1 mainly performs operations of map setting, motion detection, biological detection, status transition determination, event determination, and determination update. In addition, the processing procedure of the operation described below may be replaced, omitted, replaced, or added as appropriate as possible. Moreover, a plurality of processing procedures may be executed at the same time.

<Map settings>
As shown in FIG. 2, the living room 400 of the monitored person 9 is photographed by the infrared sensor 140 of the monitoring device 100, processed as an image, and stored as a map 133 in the storage unit 130 of the monitoring device 100.

In setting the map, first, the infrared sensor 140 measures the distance from the infrared sensor 140 to the object, and transmits the data to the control unit 120. The infrared sensor 140 is a sensor using a wide-angle infrared laser in this embodiment, but instead of the infrared sensor 140, a known means capable of measuring the distance to an object may be used.

The control unit 120 receives the distance data detected by the infrared sensor 140, processes the image, passes through the external connection units 110, 210, and 310 via the server 200, and displays the processed image on the display unit of the monitor terminal 300. Display on 350.

The observer confirms the image displayed on the display unit 350, operates the input unit 330, and specifies each area of the bed area 401, the living room area 402, the toilet area 403, and the entrance / exit area 404 in the displayed image. Then, set the map 133.

The observer defines, for example, the bed surface and the area 100 cm above the bed surface as the bed area 401. Further, the position accuracy of the bed area 401 can be ± 3 cm. In this embodiment, a specific area within the bed area 401 is further designated as the first area 411 (biological detection area). Further, the vicinity of the entrance / exit of the toilet can be designated as the toilet area 403, and the vicinity of the entrance / exit of the living room 400 can be designated as the entrance / exit area 404. The living room area 402 may be designated as the entire target area detected by the infrared sensor 140 excluding the bed area 401, the toilet area 403, and the entrance / exit area 404. Further, in this embodiment, the outdoor area is not designated on the map, and the area outside the map is set as the outdoor area. The set map 133 is transmitted from the monitor terminal 300 to the monitoring device 100 via the server 200 through the external connection units 110, 210, 310, and is stored in the storage unit 130 of the monitoring device 100 and the storage unit 240 of the server 200. Will be done.

The map is set as described above.

<Motion detection>
Next, the monitoring device 100 performs motion detection. The area and status of the monitored person are detected by motion detection.

The human body detection unit 121 of the monitoring device 100 captures a moving object based on an image taken by the infrared sensor 140 and processed by the control unit 120. The human body detection unit 121 further assumes that the center of the captured moving body is the body, and assumes that the highest point of the moving body is the head or shoulder, and extracts the feature points of the head and the body. The feature points of the head are the height, curved surface and volume, which are determined by the actual measurement at 360 °. The features of the torso are height and shoulders, chest, arms and volume, which are determined by 360 ° measurements. It is not necessary to extract all the feature points. If there is an extracted feature point, the moving object is tracked by tracking, and the moving direction (vector) is estimated using optical flow. When the moving body is stopped, the human body detection unit 121 waits while maintaining the information detected immediately before the stop, and restarts the tracking and the optical flow when the movement is resumed.

The area detection unit 121a of the human body detection unit 121 detects the area where the moving body exists based on the feature points and the vector of the moving body. When the moving body is not detected in any of the bed area 401, the living room area 402, the toilet area 403, and the entrance / exit area 404, the area detection unit 121a causes the moving body to move to the outdoor area outside the map 133. It shall be detected.

The status detection unit 121b of the human body detection unit 121 detects the status of the moving body based on the feature points and vectors of the moving body. The status of the moving body means a specific state of the moving body, and in this embodiment, the status detected by the status detection unit 121b is lying, standing, standing up, bordering, falling, crouching, or lying down. , Entering the room 400, leaving the room 400, entering (toilet), leaving (toilet). The types of statuses and the contents of each status are not limited to the following, and may be freely defined.

For example, if the highest point of the moving body in the bed area is within a height of less than 30 cm in the bed area 401, the "lying position" status is detected.

If the highest point of the moving body is within the range of 30 cm to 100 cm in height in the bed area 401, the status of "getting up" is detected. In this embodiment, in order to distinguish between turning over and getting up, the status of "getting up" is not detected if the height is less than 30 cm.

The boundary position includes three movements: sitting on the edge, crossing the fence, and sliding down. Of the boundary positions, the end sitting position is a state in which the lower body is exposed to the bedside in the sitting position, and as described above, the lower body of the moving body is outside the bed area 401, and 30% or more of the volume of the moving body is 100. If it is detected that less than% is outside the bed area 401, the status of "sitting on the edge" is detected.

Of the boundary positions, over the fence is a state in which the upper body is exposed to the bedside in the sitting state, the upper body of the moving body is outside the bed area 401, and 30% or more and less than 100% of the volume of the moving body is the bed area. When it is detected that it is outside the 401, the status of "over fence" is detected.

Of the boundary positions, sliding down is a state in which the upper or lower body is exposed to the bedside in the lying position, and it was detected that the volume of 30% or more and less than 100% of the moving body exceeds the bed frame. In that case, the status of "sliding down" is detected. If tracking is not possible by human body detection and a moving object of a certain volume is detected outside the bed area 401, it is possible that the futon has slipped off. Therefore, even in such a case, the status of "sliding down" is detected.

In the living room area 402, a "fall" status is detected when a sudden change in body position from a walking or standing state, that is, when the amount of downward displacement of the head or shoulder is equal to or more than a specified value. To. When the "fall" status is defined in this way, the "fall" status is detected even if the user has not actually fallen, but since the fall is a high-risk status if it actually occurs, In this embodiment, the status of "fall" is detected even in the near hat.

When the status of the detected monitored person changes, the status detection unit 121b stores the latest status in the storage unit 130 of the monitoring device 100 and the storage unit 240 of the server 200. If the status of the monitored person does not change, the status detection unit 121b does not have to update the status stored in the storage unit 240. For a specific status, the status detection unit 121b may store the latest status in the storage units 130 and 240 again even if the status of the detected monitored person does not change for a certain period of time.

The status detection unit 121b may display the latest status on the display unit 350 of the monitor terminal 300 via the external connection units 110, 210, 310. When the status is displayed on the display unit 350, the observer can also correct the status. For example, when the area detection unit 121a detects that the person to be monitored is in the toilet area 403 and the status detection unit 121b detects the status of entry (toilet), the status of "entry (toilet)" is displayed on the display unit 350. Is displayed. However, if the status is the result of a false positive and the monitored person actually returns from the toilet area 403 to the living room area 402, the monitor operates the input unit 330 to operate the external connection units 110 and 210. , 310 can be used to input the actual status to the control unit 120. The storage units 130 and 240 store the latest status that reflects the status corrected by the observer.

<Biological detection>
The above-mentioned motion detection may not be possible when the monitored person 9 is at rest. Therefore, when it is detected that the monitored person 9 is in the bed area 401 and no change of a predetermined size or more is detected in the first area 411 for a predetermined period, the monitoring device 1 is used. The motion detection is finished, and the biological detection for detecting the minute movement is started. The biological detection is performed by the biological detection unit 122 of the monitoring device 100.

When the monitored person is resting on the bed, the monitored person produces only minute movements. Therefore, in the method of detecting the area and the status by tracking the relatively large movement of the moving object such as the area detection unit 121a and the status detection unit 121b, the observer can use the method when the observer is resting on the bed. It may not be possible to detect if it is on the bed. On the other hand, in the detection method capable of detecting minute movements on the bed, noise is also picked up, so that noise removal becomes a problem.

Therefore, the present inventor makes use of the fact that the human body moves minutely in a certain pattern when breathing normally, and even if the absolute value of the minute movement is not detected, the minute movement We have found that the presence or absence of breathing can be detected by detecting the pattern and comparing the pattern with the pattern of movement of the human body during normal breathing.

As shown in FIG. 3A, when a person is breathing normally at rest, the body moves slightly in the same cycle as the breathing cycle. The general breathing cycle is about 4 seconds, but the breathing cycle differs depending on the individual and the physical condition. Therefore, in the biological detection of the monitoring system 1 of the present invention, the average respiratory cycle (average respiratory cycle) of the monitored subject is first measured, and then at intervals of 3/4 of the average respiratory cycle. The displacement of a specific position (first region) of the monitored person is measured. A relatively large displacement can be measured by measuring at 3/4 time intervals of the average respiratory cycle.

At the same time, at the same measurement interval, the displacement of the position (second region) where the monitored person does not exist is also measured. The displacement in the second region is noise. As shown in FIG. 3B, there is almost no displacement in the second region. Then, noise is generated by comparing the time change of the displacement of the first region and the time change of the displacement of the second region, or by the time change of the difference between the displacement of the first region and the displacement of the second region. It is possible to know whether or not the tremor due to the breathing of the monitored person is detected without having to measure the absolute value of or remove the noise.

Based on the above findings, biometric detection is performed as follows.

When the area detection unit 121a detects that the moving object exists in the bed area 401, the biometric detection unit 122 has less displacement in the bed area and is an infrared sensor than the predetermined first area 411. The area 412 located close to 140 is stored in the storage unit 130 as a second area. Further, the first measuring unit 122a of the biological detection unit 122 detects the displacement in the predetermined first region 411, and when the displacement amount becomes smaller than a predetermined size, the biological detection is started. In this embodiment, the first area 411 and the second area 412 are set in the bed area 401, but the first area 411 and the second area 412 are other areas such as the living room area 402. You may.

In the biological detection, first, the average cycle of respiration of the monitored person is measured. The first measuring unit 122a determines the average period of respiration of the monitored person from the interval between the time when the height of the first region 411 becomes the highest and the time when the height of the first region 411 becomes the lowest. Ask.

Next, the first measuring unit 122a measured the height of the first region 411 at a time interval of three-quarters of the obtained average cycle of respiration, and the second measuring unit 122b was obtained. The height of the second region 412 is measured at 3/4 time intervals of the average cycle of respiration. If the average respiration cycle cannot be detected, the first measurement unit 122a and the second measurement unit 122b may measure every 3 seconds because the general respiration cycle is 4 seconds.

The determination unit 124 determines the time change of the difference between the height in the first region 411 measured by the first measurement unit 122a and the height in the second region 412 measured by the second measurement unit 122b. It is configured to get. The determination unit 124 is further configured to determine whether or not the time change of this height difference is a predetermined pattern 134 or a change amount. If the time change of the height difference is a predetermined pattern 134 or a change amount, the determination unit 124 determines that respiration is detected in the first region 411, that is, a living body is detected. To do. If the temporal change in height difference is not a predetermined pattern 134 or change amount, the determination unit 124 determines that breathing is abnormal within the first region 411. If the time change of the difference in movement is less than a predetermined amount of change, the determination unit 124 determines that respiration is not detected, that is, the living body is not detected. If the breathing is abnormal or the breathing is not detected, it is judged to be a biological abnormality.

In this way, the time change of the movement in the first region 411 on the bed where the monitored person exists is compared with the time change of the movement in the second area 412 on the bed where the monitored person does not exist. By observing whether the difference between time-varying movements changes in a specific cycle, it is possible to more accurately detect whether or not there is a monitored person on the bed without removing noise. Can be done.

Further, as described above, in this embodiment, the movement in the first region 411 and the movement in the second region 412 are measured at intervals of 3/4 of the average respiratory cycle. By doing so, the presence or absence of respiration can be grasped with a small number of measurements.

If a relatively large movement detected by the human body detection unit 121 is detected during the biological detection, the biological detection is stopped and the motion detection is started.

<Event judgment>
The determination unit 124 collates the area, status, and the result of biological detection (whether or not it is a biological abnormality) detected at a certain detection time with the matrix 132, and sets the event of the monitored person. , It is determined whether the event is a sign of danger.

For example, when the area detection unit 121a detects that the monitored person is in the bed area 401 at 8:00 am and the status detection unit 121b detects that the status of the monitored person is rising, the rising status is set. The first status. Next, when the status detection unit 121b detects another status, the new status is set as the second status. For example, when the status detection unit 121b detects the status of the boundary position next to the status of rising, the status of the boundary position is set as the second status. Further, when the same status is detected after the lapse of a predetermined time, the status after the lapse of a predetermined time may be set as the second status. For example, if the first status is rising and the status of the monitored person remains rising even after a predetermined time has elapsed, the second status may be rising.

The determination of the event is made in two ways in this embodiment. The first method is a method of determining that a specific event always occurs if it is a combination of a specific area, first and second statuses, time, and biological detection result.

As shown in FIG. 4, for example, between 8:00 am and 9:00 am, the area detected at a certain detection time is the bed, the living body detection unit 122 detects the living body, and the first status is ". If it is "boundary position" and the second status is "getting out of bed", the determination unit 124 always determines that "getting out of bed" will occur as an event, and determines that this event is a sign of danger.

The second method is whether a particular event occurs based on a particular area, a combination of first and second stats, time of day, breathing and / or body movement, and the probability of occurrence of a past event. It is a method of judging whether or not.

As shown in FIG. 4, for example, between 9:00 and 10:00 am, the area detected at a certain detection time is the bed, the living body detection unit 122 detects the living body, and the first status is ". If the status is "boundary" and the second status is "get out of bed", the determination unit 124 will post a "toilet" event after this based on the history of past events stored in the storage unit 240 of the server 200. Calculate the probability of occurrence. If the probability is, for example, 60% or more, the determination unit 124 determines that a "toilet" event will occur, and determines that this event is a sign of danger. On the other hand, based on the history of past events, if the probability that the "toilet" event will occur after this is less than 60%, for example, the determination unit 124 does not determine that the "toilet" event will occur.

As another example of the second method, for example, between 10:00 and 11:00 am, the area detected at a certain detection time is the bed, and the living body detection unit 122 detects and detects the living body. If the first status is "get up" and the second status is "get out of bed", the determination unit 124 will follow this based on the history of past events stored by the storage unit 240 of the server 200. Calculate the probability that a "leave" event will occur. If the probability is, for example, 60% or more, the determination unit 124 determines that an event of "leaving the room" occurs, and determines that this event is a sign of danger. On the other hand, based on the history of past events, if the probability that the event of "leaving the room" will occur after this is less than 60%, for example, the determination unit 124 does not determine that the event of "leaving the room" will occur.

Which event is determined by the first method and which event is determined by the second method is arbitrary.

In this way, by predicting the occurrence of the next event based on the probability of occurrence of the past event, it is possible to predict the event more accurately according to the habits of the monitored person. In addition, the accuracy of event prediction can be maintained even when the behavioral habits and situations of the monitored person change.

For example, between 12:00 and 20:00, the biological detection unit 122 detects that there is breathing, and after the status of "lying position" is detected as the first status in the bed area 401, When the status of "entering" or "leaving" is detected as the second status, the determination unit 124 determines that the event of "entering / leaving the caregiver or family" has occurred. In this case, it is not judged to be a sign of danger. It also returns the latest status of the observer to "recumbent".

Further, for example, when the status of "lying position" is detected in the bed area 401 and the state in which breathing and / or body movement is not detected by the biological detection unit 122 continues for 2 minutes, an event of "biological abnormality" occurs. Judge that you are doing it and notify it.

When the event is determined to be a sign of danger, the control unit 120 controls the notification unit 340 so that the notification unit 340 notifies the observer.

Further, the determination unit 124 may notify the observer of the event by e-mail or short message by the communication unit that communicates with the outside via the network.

<Judgment update>
The observer notified by the notification unit 340 visually confirms the correctness of the event through the display unit 350 or directly. If the notified event is correct, the observer handles the event. On the other hand, if the notified event does not actually occur, the observer inputs to the input unit 330 that the notified event is incorrect.

The determination unit 124 stores the area, status, biological detection result, and the history of correctness of the event determined based on the time in the storage unit 240 of the server 200, and updates the probability of occurrence of the event. In this way, the observer can give feedback to the judgment by the judgment unit 124, so that the behavior of the monitored person can be estimated according to the characteristics of the watched person, the characteristics of the watched person, and changes in the environment. , It is possible to improve the accuracy of predicting the danger to the monitored person.

In addition, the observer operates the input unit 330 to give an instruction to the control unit 120 so that, for example, even if an event of "entering / leaving the caregiver or family member" occurs, the notification is not given at a specific time zone. May be good.

It is preferable that the monitoring device 100 periodically repeats the above-mentioned processing. The repetition time interval is set as appropriate. Further, the monitoring device 100 may be configured to start the monitoring process by input of the observer when the above-mentioned process is not periodically repeated.

As described above, in the monitoring system 1, the monitor gives feedback to the judgment by the judgment unit 124, so that the monitored person can respond to the characteristics of the monitored person 9, the characteristics of the monitored person 9, and changes in the environment. It is a monitoring system that can improve the accuracy of estimating the behavior of 9 and predicting the danger to the monitored person 9.

That is, the monitoring system 1 has a detection unit that detects at least one element among a plurality of elements for identifying the event of the monitoring target person 9, and a monitoring target person based on the event element detected by the detection unit. The determination unit 124 that determines the event of 9, the notification unit 340 that notifies the observer of the event determined by the determination unit 124, and the observer inputs the correctness of the event notified by the notification unit 340 to the determination unit 124. The input unit 330 is provided, and the determination of the event of the monitored person 9 by the determination unit 124 is configured to be updated based on the correctness input by the input unit 330.

In the monitoring system 1, the area detection unit 121a and the status detection unit 121b of the detection unit detect the area where the monitoring target person 9 exists and the status of the monitoring target person 9 as elements for identifying the event. The storage units 130 and 240 that store at least the combination in which the area where the monitored person 9 exists and the change in the status are associated with the event of the monitored person 9 are provided, and the determination unit 124 has at least the area detection unit 121a of the detection unit. The area detected by the status detection unit 121b and the change in status are collated with the combination stored in the storage units 130 and 240 to determine the event of the monitored person 9.

In the monitoring system 1, the detection unit is configured to further detect the presence or absence of respiration of the monitored person 9 by the biological detection unit 122, and the combination stored in the storage units 130 and 240 is the respiration of the monitored person 9. The determination unit 124 is further associated with the presence / absence of breathing, and is configured to determine the event of the monitored person 9 by collating with the combination associated with the presence / absence of respiration detected by the detection unit.

In the monitoring system 1, the detection unit is configured to further detect the time by the clock 123, the combinations stored in the storage units 130 and 240 are further associated with the time, and the determination unit 124 further detects the time. The time detected by the clock 123 is collated with the combination to determine the event of the monitored person 9.

In the monitoring system 1, the area detected by the area detection unit 121a of the detection unit includes the bed area 401, the living room area 402, the toilet area 403, the entrance / exit area 404, and the outside of the living room.

In the monitoring system 1, the status of the monitored person 9 includes lying down, standing, getting up, boundary position, getting out of bed, entering and leaving the room, falling, crouching, and lying down.

The monitoring method executed in the monitoring system 1 of the present embodiment includes a step in which the monitoring device 100 detects at least one element among a plurality of elements for identifying an event of the monitored person 9. Based on the step of determining the event of the monitored person 9 based on the event element, the step of notifying the observer of the determined event, and the correctness of the event notified to the observer input by the observer. It includes a step of updating the event judgment of the monitored person 9.

The monitoring program executed in the monitoring system 1 of the present embodiment is based on a step of detecting at least one element of a plurality of elements for identifying an event of the monitored person 9 and an element of the detected event. Based on the step of determining the event of the monitored person 9 and the step of notifying the monitor of the determined event, and the correctness of the event notified to the monitor input by the monitor, the monitored person 9 This is a monitoring program for causing the monitoring device 100 to execute a step of updating the determination of an event.

The recording medium of the present embodiment is a recording medium on which the above-mentioned program is recorded.

Further, the monitoring system 1 improves the accuracy of predicting the danger to the monitored person 9 by determining the event of the monitored person 9 based on not only the status of the monitored person 9 but also the status and the area. It is a monitoring system that can be used.

That is, the monitoring system 1 has an area detection unit 121a and a status detection unit 121b that detect the area where the monitoring target person 9 exists and the status of the monitoring target person 9 as elements for identifying the event of the monitoring target person 9. The monitoring unit 9 includes a detection unit including, a determination unit 124 that determines the event of the monitoring target person 9 based on the event element detected by the detection unit, and at least the area and status in which the monitoring target person 9 exists. The storage units 130 and 240 are provided for storing the combinations associated with the events of the above, and the determination unit 124 collates at least the area detected by the detection unit and the change in the status with the combinations stored in the storage units 130 and 240. , It is configured to judge the event of the monitored person 9.

In the monitoring system 1, the detection unit is configured to further detect the presence or absence of respiration of the monitored person 9 by the biological detection unit 122, and the combination stored in the storage units 130 and 240 is the monitoring target person 9. Further associated with the presence or absence of respiration, the determination unit 124 is configured to determine the event of the monitored person 9 by further collating with the combination associated with the presence or absence of respiration detected by the detection unit.

In the monitoring system 1, the detection unit is configured to further detect the time by the clock 123, the combinations stored in the storage units 130 and 240 are further associated with the time, and the determination unit 124 is further by the detection unit. The event of the monitored person 9 is determined by collating the detected time with the combination.

In the monitoring system 1, the area detected by the detection unit includes the bed area 401, the living room area 402, the toilet area 403, the entrance / exit area 404, and the outside of the living room.

In the monitoring system 1, the status of the monitored person 9 includes lying down, standing, getting up, boundary position, getting out of bed, entering and leaving the room, falling, crouching, and lying down.

In addition, the monitoring system 1 includes a notification unit 340 that notifies the observer of an event determined by the determination unit 124.

The monitoring system 1 includes an input unit 330 in which the monitor inputs the correctness of the event notified by the notification unit 340 to the determination unit 124, and the determination of the event of the monitored person 9 by the determination unit 124 is input by the input unit 330. It is configured to be updated based on the correctness made.

In the monitoring method executed in the monitoring system 1 of the present embodiment, as an element for the monitoring device 100 to identify the event of the monitored person 9, the area where the monitored person 9 exists and the status of the monitored person 9 are present. The step of detecting the above, the step of determining the event of the monitored person 9 based on the detected event element, and at least the area and status in which the monitored person 9 exists are associated with the event of the monitored person 9. It includes a step of storing the combination and a step of collating at least the detected area and status with the stored combination to determine the event of the monitored person 9.

The monitoring program executed in the monitoring system 1 of the present embodiment is a step of detecting the area where the monitoring target person 9 exists and the status of the monitoring target person 9 as an element for identifying the event of the monitoring target person 9. And the step of determining the event of the monitored person 9 based on the detected event element, and the step of storing at least the combination of the area and status in which the monitored person 9 exists and the event of the monitored person 9 are stored. This is a monitoring program for causing the monitoring device 100 to execute a step of collating at least the detected area and the status with the stored combination and determining the event of the monitored person 9.

The recording medium of the present embodiment is a recording medium on which the above-mentioned program is recorded.

In addition, the monitoring system 1 can detect the presence or absence of breathing without detecting the absolute value of minute movements when there is almost no movement, such as when the monitored person 9 is in a resting state. It is a system.

That is, in the monitoring system 1, the first measuring unit 122a for measuring the movement in the first area 411 in which the monitored person 9 exists, and the first measuring unit 122a, which is different from the first area 411, in which the monitored person 9 does not exist. The second measuring unit 122b that measures the movement in the second region 412, the movement in the first region 411 measured by the first measuring unit 122a, and the second measuring unit 122b measured by the second measuring unit 122b. The control unit 120 is provided with a control unit 120 configured to acquire a time change of the difference from the movement in the region 412, and the control unit 120 further indicates whether the time change of the movement difference is a predetermined pattern or change amount. It is configured to determine whether or not.

Further, in the monitoring system 1, the movement in the first region 411 and the movement in the second region 412 are measured at a time interval of 3/4 of the average respiratory cycle.

Further, in the monitoring system 1, the first measuring unit 122a and the second measuring unit 122b are configured to measure minute movements in a non-contact state using an infrared laser.

Further, the monitoring system 1 includes a notification unit 340 that notifies the observer when the difference in movement is equal to or less than a predetermined value for a predetermined time.

Further, in the monitoring system 1, the first area 411 is an area on the bed area 401.

Further, the monitoring method executed in the monitoring system 1 of the present embodiment is different from the step in which the monitoring device 100 measures the movement in the first area 411 in which the monitored person 9 exists and the first area 411. , The time change of the difference between the step in which the monitoring device 100 measures the movement in the second region 412 in which the monitoring target person 9 does not exist and the movement in the first region 411 and the movement in the second region 412. The step includes a step acquired by the monitoring device 100 and a step in which the monitoring device 100 determines whether or not the time change of the movement difference is a predetermined pattern or change amount.

Further, the monitoring program executed in the monitoring system 1 of the present embodiment has a step of measuring the movement in the first area 411 in which the monitoring target person 9 exists, and a monitoring target person different from the first area 411. The step of measuring the movement in the second region 412 in which 9 does not exist, the step of acquiring the time change of the difference between the movement in the first region 411 and the movement in the second region 412, and the difference in movement. This is a monitoring program for causing the monitoring device 100 to execute a step of determining whether or not the time change of the above is a predetermined pattern or change amount.

Further, the recording medium of the present embodiment is a recording medium on which the above-mentioned program is recorded.

It should be considered that the embodiments disclosed above are exemplary in all respects and not restrictive. The scope of the present invention is not shown in the above embodiments, but is indicated by the claims and includes all modifications and modifications within the meaning and scope equivalent to the claims.

1: Monitoring system, 100: Monitoring device, 120: Control unit, 121: Human body detection unit, 121a: Area detection unit, 121b: Status detection unit, 122: Biometric detection unit, 122a: First measurement unit, 122b: First 2 measurement unit, 123: clock, 124: judgment unit, 130: storage unit, 131: program, 132: matrix, 134: pattern, 140: infrared sensor, 240: storage unit, 330: input unit, 340: notification unit , 350: Display, 401: Bed area, 402: Living room area, 403: Toilet area, 404: Doorway area, 411: First area, 412: Second area.

Claims (14)

As elements for identifying the event of the monitored person, a detector that detects the time, the area where the monitored person exists, and the change in the status of the monitored person,
A judgment unit that determines the event of the monitored person based on the event element detected by the detection unit,
For each plurality of times, and a storage unit that stores combinations associated with monitored subject events and an area and a plurality of status in the presence of at least the monitored person,
The determination unit is configured to determine an event of the monitored person by collating at least the time, area, and status change detected by the detection unit with the combination stored in the storage unit. There is a monitoring system. The combination stored in the storage unit is set for each of a plurality of time zones.
The monitoring system according to claim 1, wherein the determination unit is configured to further collate the time detected by the detection unit with the combination to determine the event of the monitoring target person. The determination unit collates at least the time, area, and status change detected by the detection unit with the combination stored in the storage unit, and is based on the probability of occurrence of a past event of the monitored person. The monitoring system according to claim 1 or 2, which is configured to determine the event of the monitored person. The detection unit is further configured to detect the presence or absence of breathing of the monitored person.
The combination stored in the storage unit is further associated with the presence or absence of breathing of the monitored person.
Any of claims 1 to 3, wherein the determination unit is configured to determine an event of the monitored person by further collating with the combination associated with the presence or absence of respiration detected by the detection unit. monitoring system according to any. The monitoring system according to any one of claims 1 to 4, wherein the area detected by the detection unit includes a bed, a living room, a toilet, an entrance / exit, and an outside of the living room. The monitoring system according to any one of claims 1 to 5, wherein the status of the monitored person includes lying down, standing, getting up, boundary position, getting out of bed, entering and leaving the room, falling, crouching, and lying down. The monitoring system according to any one of claims 1 to 6 , further comprising a notification unit that notifies the observer of an event determined by the determination unit. It is provided with an input unit in which the observer inputs the correctness of the event notified by the notification unit to the determination unit.
The monitoring system according to claim 7 , wherein the determination of the event of the monitored person by the determination unit is configured to be updated based on the correctness input by the input unit. Information processing device
As elements for identifying the event of the monitored person, a step of detecting the time, the area where the monitored person exists, and the change in the status of the monitored person, and
Steps to determine the event of the monitored person based on the detected event element ,
Including
In the step of determining the event, at least the detected time, area, and status are stored in association with the event of the monitored person, and at least the area where the monitored person exists and the plurality of statuses are stored at multiple times. against the pair combined, it determines the person to be monitored for events, monitoring method. The combination is set for each of a plurality of time zones.
The monitoring method according to claim 9, wherein the step of determining the event is configured to match the detected time with the combination to determine the event of the monitored person. The step of determining the event collates the detected time, the area, and the change in status with the stored combination, and the monitored person is based on the probability of occurrence of a past event of the monitored person. The monitoring method according to claim 9 or 10, which is configured to determine the event of. For information processing equipment
As elements for identifying the event of the monitored person, a step of detecting the time, the area where the monitored person exists, and the change in the status of the monitored person, and
Steps to determine the event of the monitored person based on the detected event element,
It is a monitoring program to realize
In the step of determining the event, at least the detected time, area, and status are stored in association with the event of the monitored person, and at least the area where the monitored person exists and the plurality of statuses are stored at a plurality of times . against the set together, determines the person to be monitored for events,
A monitoring program to make that happen . The combination is set for each of a plurality of time zones.
The monitoring program according to claim 12, wherein the step of determining the event is configured to match the detected time with the combination to determine the event of the monitored person. The step of determining the event collates the detected time, the area, and the change in status with the stored combination, and the monitored person is based on the probability of occurrence of a past event of the monitored person. The program according to claim 12 or 13, which is configured to determine the event of.

Images