JP2021140756A - Method of watching, system, program, recording medium, device, and dwelling - Google Patents

Abstract

To provide a watching method for acquiring behavior pattern information from life sound information of a watching object person to detect a behavior abnormality by targeting a private space such as a bathroom, a dressing room, a restroom, and a bedroom.SOLUTION: A watching system 2 comprises steps of: acquiring watching area information by considering any of a bathroom 6-1, a dressing room 6-2, a restroom 6-3, and a bedroom 6-4, or a private space 6 including two or more of them as a watching area 4; acquiring watching object person information; acquiring life sound information of the watching object person; acquiring event information representing a behavior of the watching object person from the life sound information; acquiring behavior pattern information representing a behavior pattern associated with the watching area information and the watching object person information from the acquired event information; and monitoring coincidence/non-coincidence between the behavior pattern information and the latest event information, monitoring duration time of a non-coincident state, and determining there is an abnormality in the watching object person when the duration time is a threshold or more.SELECTED DRAWING: Figure 2

Description

The present disclosure relates to a monitoring technique that monitors behavior by living sounds emitted in a private space of a bathroom, a dressing room, a toilet, and a bedroom, and detects an abnormality of a person to be monitored.

Aside from the case where a caregiver accompanies a person's life, a free society is usually constructed in which the privacy of the person is respected and a free life is ensured. There is a limit to the constant monitoring of elderly people, care recipients, etc. by others and their families, and if the watching is strict, the psychological resistance to the watching target will be amplified.
For such a watching system, a sound sensor device is installed in the living space, the sound generated in the living space is collected, the controller generates a simplified spectrogram in both time and frequency, and the watching server filters the spectrogram. It is disclosed that processing is performed to calculate a score value, and the living space and the situation of the subject are recognized from this score value (for example, Patent Document 1).

When grasping the situation of the person to be watched, replacing the first situation with the first sound, replacing the second situation with the second sound, and notifying using the first sound. It is known that the notification by the second sound reduces the feeling of psychological resistance to notify the situation in the real space (for example, Patent Document 2).
Regarding the information used for watching, it is disclosed that life sound information is collected, this life sound information is analyzed, and the analysis result is notified to a guardian (for example, Patent Document 3).
Regarding the notification of an abnormality report, it is disclosed that a living sound is detected, compared with the abnormal sound, and the abnormality information is notified (for example, Patent Document 4).
Regarding the detection of abnormal operation, it is disclosed that a specific frequency is extracted from the detection proposal to detect the abnormal operation state (for example, Patent Document 5).

Japanese Unexamined Patent Publication No. 2011-237856 Japanese Unexamined Patent Publication No. 2019-175371 Japanese Unexamined Patent Publication No. 2013-210799 Japanese Unexamined Patent Publication No. 2008-197879 Japanese Unexamined Patent Publication No. 2004-06291

For example, it is possible to acquire the living sounds of infants and the elderly as a person to be watched over, but it is not easy to detect an abnormality from the living sounds. In particular, in sound information, the existence of noise components other than the element of abnormality determination cannot be ignored. This is because the noise component causes erroneous judgment of abnormality judgment and judgment error.
Living sound information includes various miscellaneous sounds, and it is not easy to grasp the living sound information from these sound collection information, and even if the living information is collected, it is considerably difficult to judge. Misjudgment cannot be denied. It is necessary for the administrator to confirm it, and if the confirmation frequency is high, there is a problem that it lacks practicality.
Regarding such a monitoring system, new construction of a house is an opportunity to install the system, but a large-scale system that is premised on renovation is not realistic due to difficulties such as introduction cost.

By the way, there is a certain degree of regularity in the lives and behaviors of protected persons and the elderly, and the behaviors tend to become habitual. Especially, the tendency is remarkable in the life in the facility.
The inventor has found that it is effective to pattern and watch the behavior of the person being watched over by using the living sounds emitted from, for example, getting up, washing, eating, entertainment, reading, bathing, and going to bed.
Therefore, in view of the above problems and the above findings, the purpose of the present disclosure is to target a private space including any one of a bathroom, a dressing room, a toilet, and a bedroom, or two or more of them, and behave from the living sound information of the person to be watched over. The purpose is to acquire pattern information and detect behavioral abnormalities.

In order to achieve the above object, according to one aspect of the monitoring method of the present disclosure, it is a monitoring method using a computer, and is private including any one of a bathroom, a dressing room, a toilet, a bedroom, or two or more of them. The space is used as a watching area, and the watching area information acquisition unit acquires the watching area information representing the watching area, and the watching target person information acquisition unit acquires the watching target person information representing the watching target person. The process in which the living sound information acquisition unit collects the living sounds of the watching target person in the private space and acquires the living sound information, and the event information acquisition unit obtains the behavior of the watching target person from the living sound information. The process of detecting and acquiring the event information representing the behavior and the behavior pattern information acquisition unit collect the event information to generate the behavior pattern of the watching target person, and generate the watching area information and the watching target person information. The process of acquiring the behavior pattern information representing the behavior pattern related to the above, and the abnormality determination unit compares the behavior pattern information with the latest event information, monitors the match / mismatch between the two, and the mismatch state. If at least the duration is equal to or longer than the threshold time, the step of monitoring the duration of the above and determining that the person to be watched is abnormal is included.

In this monitoring method, the abnormality determination unit further sets a non-steady state if the latest event information includes event information that matches the event information existing in the behavior pattern information, and a steady state if there is no event information. When the duration of the steady state exceeds the threshold time, or when the behavior pattern information includes excessive sound information that does not exist, the step of determining the watching target person abnormality may be included.
In this monitoring method, the behavior pattern information unit further accumulates the event information for a predetermined period including a day, a week, a month, and a year to generate the behavior pattern information, or updates the event information. It may include a step of generating behavior pattern information.
In this monitoring method, further, when the abnormality determination unit determines the abnormality of the monitoring target person, there is no step of the reporting unit calling to the location of the monitoring target person to output voice information and no response from the monitoring target person. When the duration from the output of the calling voice exceeds the threshold time, the reporting unit may include a step of outputting alarm information.
In this monitoring method, the event information acquisition unit may further include a step of acquiring the event information by using at least one or more of the sound pressure, the frequency, and the period of the living sound information.

In order to achieve the above object, according to one aspect of the watching system of the present disclosure, a private space including any one of a bathroom, a dressing room, a toilet, a bedroom, or two or more of them is set as a watching area, and the watching area is designated as a watching area. The watching area information acquisition unit that acquires the watching area information to be represented, the watching target person information acquisition unit that acquires the watching target person information that represents the watching target person, and the living sound of the watching target person in the private space are collected. The event information is assembled by the living sound information acquisition unit that acquires the living sound information, the event information acquisition unit that detects the behavior of the watching target person from the living sound information, and acquires the event information representing the behavior. A behavior pattern information acquisition unit that generates a behavior pattern of the watching target person and acquires behavior pattern information representing the behavior pattern associated with the watching area information and the watching target person information, and the latest behavior pattern information. By comparing with the above-mentioned event information, the match and disagreement between the two are monitored, and the duration of the disagreement state is monitored. Be prepared.

In this monitoring system, the abnormality determination unit further sets a non-steady state if the latest event information includes event information that matches the event information existing in the behavior pattern information, and a steady state if there is no event information. When the duration of the steady state exceeds the threshold time, or when the behavior pattern information includes excessive sound information that does not exist, the watching target person abnormality may be determined.
In this monitoring system, the behavior pattern information unit may further accumulate the event information in a predetermined period including a day, a week, a month, and a year, or update the event information to generate the behavior pattern information. ..
In this monitoring system, when the abnormality determination unit determines the abnormality, the call voice information is generated at the location of the watch target person, and there is no response from the watch target person, and the call voice information is output. When the duration of the above exceeds the reference time, alarm information indicating the abnormality may be generated.
In this monitoring system, the behavior pattern information unit may further acquire the event information from the living sound information using at least one or more of sound pressure, frequency, and period.

In order to achieve the above object, according to one aspect of the program of the present disclosure, it is a program to be executed by a computer, and is private including one of a bathroom, a dressing room, a toilet, a bedroom, or two or more of them. The space is used as a watching area, and the function of acquiring the watching area information representing the watching area, the function of acquiring the watching target person information representing the watching target person, and the living sound of the watching target person in the private space are collected. , The function of acquiring the living sound information, the function of detecting the behavior of the watching target person from the living sound information, and the function of acquiring the event information representing the behavior, and the behavior of the watching target person by collecting the event information. A function of generating a pattern and acquiring behavior pattern information representing the behavior pattern associated with the monitoring area information and the monitoring target person information, and comparing the behavior pattern information with the latest event information, both of them. The computer is made to execute a function of monitoring the matching and disagreement of the above and monitoring the duration of the disagreement state, and determining that the monitoring target person is abnormal if the duration is at least the threshold time or more.

In this program, if the latest event information includes event information that matches the event information existing in the behavior pattern information, it is in a non-steady state, and if there is no event information, it is in a steady state, and the duration of the steady state is a threshold. When the time is exceeded, or when the behavior pattern information includes excessive sound information that does not exist, the computer may be made to execute the function of determining the abnormality of the watching target person.
In this program, a function of accumulating the event information for a predetermined period including a day, a week, a month, and a year to generate the behavior pattern information, or updating the event information to generate the behavior pattern information is provided. You may let the computer do it.
In this program, further, when the abnormality of the watching target person is determined, the reporting unit outputs a call voice information to the location of the watching target person, and there is no response from the watching target person and the call voice is output. When the duration of the above exceeds the threshold time, the computer may be made to execute the function of outputting the alarm information by the reporting unit.
In this program, the computer may further perform a function of acquiring the event information by using at least one or more of the sound pressure, the frequency, and the period of the living sound information.

In order to achieve the above object, according to one aspect of the recording medium of the present disclosure, it is a recording medium in which the program is stored.

In order to achieve the above object, according to one aspect of the watching device of the present disclosure, it is a watching device that serves as a watching area in a private space including any one of a bathroom, a dressing room, a toilet, and a bedroom, or two or more of them. The device unit is provided with a watching area information acquisition unit that acquires watching area information representing the watching area, a watching target person information acquisition unit that acquires watching target person information representing a watching target person, and the like. The living sound information acquisition unit that collects the living sounds of the watching target person in the private space and acquires the living sound information, and the event information that detects the behavior of the watching target person from the living sound information and represents the behavior. The event information acquisition unit that acquires The behavior pattern information acquisition unit that acquires If so, it is equipped with an abnormality determination unit that determines that the person being watched over is abnormal.

In this monitoring device, the abnormality determination unit further sets a non-steady state if the latest event information includes event information that matches the event information existing in the behavior pattern information, and a steady state if there is no event information. When the duration of the steady state exceeds the threshold time, or when the behavior pattern information includes excessive sound information that does not exist, the watching target person abnormality may be determined.
In this monitoring device, the behavior pattern information unit may further accumulate the event information or update the event information in a predetermined period including a day, a week, a month, and a year.
In this monitoring device, when the abnormality determination unit determines the abnormality, it generates call voice information at the location of the watch target person, and there is no response from the watch target person, and the call voice is output from the output of the call voice. When the duration exceeds the reference time, alarm information indicating the abnormality may be generated.
In this monitoring device, the event information acquisition unit may further acquire the event information from the living sound information using at least one or more of sound pressure, frequency, and period.

In order to achieve the above object, according to one aspect of the dwelling of the present disclosure, the dwelling is equipped with the watching method, the watching system, the program, the recording medium, and the watching device.

According to the present disclosure, any of the following effects can be obtained.
(1) A private space including one of the bathroom, dressing room, toilet, and bedroom, or two or more of these, can be watched as a watching area, and the person to be watched can be watched. The person's abnormality can be detected.
(2) The behavior pattern information acquired from the watching target person in the specified watching area is compared with the latest event information, an abnormality is detected due to a match mismatch, and the duration of this abnormal state continues beyond the threshold time. When this happens, it can be determined that the person being watched over is abnormal.
(3) Abnormality of the person being watched is determined using the living sound pattern information generated in the bathroom, dressing room, toilet, and bedroom of the person being watched, and alarm information is output based on the continuation of the abnormality, thus avoiding the continuation of the abnormal state. It can protect the target person from unforeseen circumstances.
(4) As an existing private space, it can be installed in the bathroom, dressing room, toilet, and bedroom to easily build a watching system for the person to be watched.

It is a flowchart which shows the monitoring process which concerns on one Embodiment. It is a figure which shows the watching system which concerns on one Embodiment. It is a figure which shows a watching area and a living sound source. It is a figure which shows the bathroom, the dressing room, the toilet and the bedroom as the place of the person to be watched over. It is a figure which shows the watching system which concerns on one Embodiment. It is a figure which shows an example of the hardware of a watching unit. It is a flowchart which shows the acquisition process of behavior pattern information. It is a figure which shows the acquisition process of the behavior pattern information. It is a flowchart which shows the processing procedure of anomaly detection logic. It is a figure which shows the abnormality detection logic processing example 1. FIG. It is a figure which shows the abnormality detection logic processing example 2. It is a figure which shows the abnormality detection logic table. It is a figure which shows an example of a monitoring management database. It is a figure which shows an example of the behavior pattern information database. It is a figure which shows an example of the event / behavior pattern comparison database. It is a figure which shows the modification of the watching system. It is a figure which shows the other modification of the watching system. It is a figure which shows an example of the watching process which concerns on one Example. It is a flowchart which shows the monitoring unit management process. It is a flowchart which shows the abnormality monitoring / calling process. It is a flowchart which shows the location determination process. It is a figure which shows an example of the hardware of the monitoring unit which concerns on one Example. It is a figure which shows the installation example of the monitoring unit which concerns on one Example.

[One Embodiment]
FIG. 1 shows a processing procedure of a monitoring process according to an embodiment. The procedure shown in FIG. 1 is an example, and the present disclosure is not limited to the procedure shown in FIG.
This watching process is an example of a watching method using a computer, and in this watching process, acquisition of watching area information (S101), acquisition of watching target person information (S102), acquisition of living sound information (S103), and watching. Acquisition of the location information of the target person P (S104), acquisition of event information (S105), acquisition of behavior pattern information (S106), comparison of the latest event information and behavior pattern information (S107, S108), duration of steady state Measurement (S109), duration determination (S110), excessive sound information determination (S111), monitoring target person abnormality determination (S112), call voice information generation and output (S113), response determination (S114), It includes determination of the duration of an abnormality (S115), generation and output of alarm information (S116), and the like.

Acquisition of watching area information (S101): The watching area information acquisition unit 18 acquires watching area information that identifies the watching area 4 of the watching target person P. This watching area information is address information for specifying the watching area of the watching target person P. This monitoring area information is registered in the monitoring management database (DB) 52 (FIG. 13) by the processing unit 10.
Acquisition of watching target person information (S102): The watching target person P to be watched in the watching area 4 is specified, and the watching target person information acquisition unit 20 inputs the name, identification information, etc. of the watching target person P to the information input unit 14. Acquires the information of the person to be watched over. This monitoring target person information is registered in the monitoring management DB 52. As a result, the watch target person P is related to the watch area information.

Acquisition of living sound information (S103): This living sound information includes living sounds related to the watching target person P and environmental sound information generated from the whereabouts L of the watching target person P. The living sound of the person to be watched P is collected by the microphone 36 installed in the place L. This sound collection signal is input to the processing unit 10, and the living sound information acquisition unit 22 acquires the living sound information.
Acquisition of the location information of the monitoring target person P (S104): The location L of the monitoring target person P can be specified from the living sound information representing the living sound emitted from the location L. This living sound information includes movement information of the person P to be watched over, and the location L can be specified from this movement information as well. Then, the location information representing the specified location L is registered in the monitoring management DB 52 in relation to the monitoring target person information.

Acquisition of event information (S105): Acquires event information representing an event related to the watching target person P from the living sound information acquired from the location L of the watching target person P. This event is, for example, an acoustic change due to a state transition from a constant operating state to another operating state, and this change information may be used as event information. For the acquisition period of this event information, for example, a collection period such as a daily unit, a multi-day unit, a monthly unit, or a year unit is set, and the event information is continuously acquired and accumulated. This event information is accumulated in the monitoring management DB 52.
Acquisition of behavior pattern information (S106): The behavior pattern information is generated by aggregating a plurality of event information obtained during the accumulation period, and is behavior information representing the behavior of the watch target person P in the life. This behavior pattern information is registered in the behavior pattern information DB 74 (FIG. 14) in relation to the monitoring target person information.

Comparison of the latest event information and behavior pattern information (S107, S108): The abnormality determination unit 28 has the latest event information acquired by the event information acquisition unit 24 and the behavior pattern information acquired by the behavior pattern information acquisition unit 25. To determine whether the latest event information matches the behavior pattern information. That is, in this comparison, if the latest event information and the behavior pattern information match, it is determined to be a non-steady state (normal), and if they do not match, it is determined to be a steady state (abnormal). This determination process and comparison result information are registered in the event / behavior pattern comparison DB88 (FIG. 15).
Measurement of Steady State Duration (S109): When the abnormality determination unit 28 determines the steady state, the duration is measured starting from the occurrence of this steady state. That is, this duration is the length of time that the steady state lasts. A threshold time is set as a criterion for this duration.

Determination of duration (S110): The abnormality determination unit 28 determines whether or not the duration of the steady state exceeds the threshold time.
Determination of excessive sound information (S111): When the duration of the steady state does not exceed the threshold time (NO in S110), the abnormality determination unit 28 determines whether the latest event information includes excessive sound information. , Measure the duration of steady state.
Determining Abnormality of Watching Target Person (S112): When the duration of the steady state exceeds the threshold time (YES in S110), or when the latest event information includes excessive sound information (YES in S111). ), Abnormal (= continuation of steady state).
Generation and output of call voice information (S113): When the abnormality determination unit 28 determines an abnormality, the processing unit 10 generates call voice information and outputs the call voice information.

Response determination (S114): The abnormality determination unit 28 determines whether there is a response from the watching target person P, and confirms the abnormality of the watching target person P. The response voice from the monitoring target person P is collected by the microphone 36, and the abnormality determination unit 28 monitors the presence or absence of the response of the watching target person P from the sound collection information. If the response information representing the response voice is acquired (YES in S114), the process returns to S105.
Determining the duration of the abnormality (S115): If there is no response from the watching target person P (NO in S114), the abnormality determination unit 28 determines whether the duration from the output time of the calling voice information exceeds the threshold time. do.
Generation and output of alarm information (S116): When the duration of the steady state exceeds the threshold time (YES in S115), the abnormality determination unit 28 generates and outputs alarm information. This history information is recorded in the monitoring management DB 52 and the event / behavior pattern comparison DB 88.

<Watching system 2>
FIG. 2 shows an example of the monitoring system 2 according to the embodiment. In this watching system 2, a private space 6 is set as a watching area 4 of the watching target person P. The private space 6 is a space in consideration of the privacy of the person P to be watched over, and includes any one of the bathroom 6-1 and the dressing room 6-2, the toilet 6-3, and the bedroom 6-4, or two or more of them. .. That is, the watching area 4 includes a private space 6 including one or more selected from the bathroom 6-1, the dressing room 6-2, the toilet 6-3, and the bedroom 6-4. Therefore, in the private space 6,
a) Bathroom 6-1
b) Dressing room 6-2
c) Toilet 6-3
d) Bedroom 6-4
e) Bathroom 6-1 and dressing room 6-2
f) Bathroom 6-1 and toilet 6-3
g) Bathroom 6-1 and bedroom 6-4
h) Dressing room 6-2 and toilet 6-3
i) Dressing room 6-2 and bedroom 6-4
j) Toilet 6-3 and bedroom 6-4
k) Bathroom 6-1, dressing room 6-2 and toilet 6-3
l) Bathroom 6-1, dressing room 6-2 and bedroom 6-4
m) Bathroom 6-1 and toilet 6-3 and bedroom 6-4
n) Dressing room 6-2, toilet 6-3 and bedroom 6-4
o) Bathroom 6-1, dressing room 6-2, toilet 6-3 and bedroom 6-4
Etc. are included.

The monitoring system 2 includes a sound collecting unit 8, a processing unit 10, a communication unit 12, an information input unit 14, and an information presenting unit 16.
The sound collecting unit 8 collects sound from at least the watching target person P in the watching area 4, and provides the sound collecting information to the processing unit 10.
The processing unit 10 includes a watching area information acquisition unit 18, a monitoring target person information acquisition unit 20, a living sound information acquisition unit 22, an event information acquisition unit 24, a behavior pattern information acquisition unit 25, and an abnormality determination unit 28.

The watching area information acquisition unit 18 acquires the watching area information representing the watching area 4. Watching area information representing the watching area 4, such as identification information for identifying whether the watching area 4 is one or more of the bathroom 6-1, the dressing room 6-2, the toilet 6-3, and the bedroom 6-4. Is obtained from the information input unit 14.
The monitoring target person information acquisition unit 20 acquires the monitoring target person information. The information of the person to be watched, such as the name, age, and image of the person to be watched P, is input from the information input unit 14.
The living sound information acquisition unit 22 acquires the living sound information of the watching target person P from the sound collecting information detected by the sound collecting unit 8 and acquired by the processing unit 10. This living sound information is acquired in relation to the private space information representing the bathroom 6-1, the dressing room 6-2, the toilet 6-3, and the bedroom 6-4.

The event information acquisition unit 24 acquires event information representing an event such as the operation of the watching target person P from the sound collection information acquired by the processing unit 10. This event information may be acquired using at least one or more of the sound pressure, frequency, and period included in the sound collection information.
By accumulating the event information, the behavior pattern information acquisition unit 25 acquires the behavior pattern information representing the behavior of the watching target person P from the event information group.
The abnormality determination unit 28 compares the event information acquired by the event information acquisition unit 24 with the behavior pattern information acquired by the behavior pattern information acquisition unit 25, and determines whether the monitoring target person P is at least in a non-steady state or in a steady state. It is determined whether the information is true, and the abnormality of the person to be watched over is determined. For information processing of the abnormality determination unit 28,
a) Comparison between behavior pattern information and the latest event information b) Monitoring of the duration when the behavior pattern information and the latest event information are inconsistent c) Set a threshold time for this duration and continue It is determined whether the time is equal to or longer than the threshold time (duration ≥ threshold time), and when the duration ≥ threshold time, the determination of the watching target person abnormality d) the judgment of the watching target person abnormality by acquiring the excessive sound information, etc. included.

The communication unit 12 communicates with a functional unit in the system and communicates with an external device. For example, it is transmitted to the communication unit 12 and the communication terminal of the administrator to make a report.
The information input unit 14 is used, for example, by the administrator to input information such as area information representing a watching area and watching target person information representing a watching target person P by using a user terminal such as a keyboard. The information input unit 14 is used for various types of information input.
When the abnormality determination unit 28 determines the abnormality of the watching target person, the information presenting unit 16 presents alert information to the location of the watching target person P. The information presenting unit 16 is provided with a reporting unit 26. When the abnormality determination unit 28 determines the monitoring target abnormality, the reporting unit 26 reports the monitoring target person abnormality to the administrator of the watching target person P.
When the abnormality determination unit 28 determines that the monitoring target person is abnormal, the reporting unit 26 outputs a call sound to the location S of the monitoring target person P. Then, when there is no response from the watching target person P and the duration from the output of the calling voice exceeds the threshold time, the reporting unit 26 outputs warning information indicating an abnormality of the watching target person. The warning information from the reporting unit 26 may be reported to the communication terminal using the communication unit 12.

<Watching area 4>
FIG. 3 shows an example of the watching area 4. This watching area 4 is an example of the residence of the present disclosure. The watching area 4 includes a bathroom 6-1, a dressing room 6-2, a toilet 6-3, and a bedroom 6-4.
The bathroom 6-1 has a door 30-1, the dressing room 6-2 has a door 30-2, the toilet 6-3 has a door 30-3, and the bedroom 6-4 has a door 30-4.
In this watching area 4, doors 30-1, 30-2, 30-3, 30-4, a person to be watched P, an air conditioner, a ventilation fan, a water tap, a shower tap, a hot water tap, a mist sauna device, etc., which are not shown, are used as sound sources. Includes a number of devices that produce sound.
In this watching area 4, for example, as shown by the broken line arrow, the watching target person P goes from the aisle to the bedroom 6-4, from the dressing room 6-2 to the toilet 6-3, and from the dressing room 6-2 to the bathroom 6-1. Includes flow lines such as moving.

<Whereabouts L of watching area 4 and watching target person P>
FIG. 4 shows the relationship between the watching area 4 and the location L. The watching area 4 is divided into a plurality of locations L1, L2, L3, and L4. In this example, bathroom 6-1 = whereabouts L1, dressing room 6-2 = whereabouts L2, toilet 6-3 = whereabouts L3, bedroom 6-4 = whereabouts L4. When the location is not specified for the location L1, L2, L3, and L4, it is simply referred to as the location L.
At the locations L1, L2, L3, and L4, watching units 32-1, 32-2, 32-3, and 32-4 are arranged to collect living sounds and detect the behavior of the watching target person P. ..

<Watching system 2>
FIG. 5 shows an example of the watching system 2. The monitoring system 2 constitutes a network including a plurality of monitoring units 32-1, 32-2, 32-3, and 32-4. When the location of the watching unit 32-1, 32-2, 32-3, 32-4 is not specified, it is simply referred to as the watching unit 32.
Under the control of the processing unit 10, the communication unit 12 links the other monitoring units 32-2, 32-3, and 32-4 via the network 34 by wireless connection. The communication unit 12 of the other monitoring units 32-2, 32-3, and 32-4 also has the same function.
As shown in the watching unit 32-1, the watching unit 32 includes at least a processing unit 10, a communication unit 12, a microphone 36, a speaker 38, and the like. That is, the processing unit 10 may be configured by a computer having a communication function.
The microphone 36 collects sound from the location L of the monitoring target person P and provides the sound collection signal to the processing unit 10.
The speaker 38 is controlled by the processing unit 10 and outputs a voice such as a calling voice to the location L.
The monitoring system 2 is composed of four monitoring units 32-1, 32-2, 32-3, and 32-4, but may be less than four or more than four.

<Information processing by the processing unit 10 of each monitoring unit 32>
For information processing by the processing unit 10,
a) Acquisition of sound collection information from location L1, L2, L3, L4 b) Identification of monitoring target person P, location L1, L2, L3, L4 c) Behavior monitoring of monitoring target person P d) Event from sound collection information Acquisition of information e) Acquisition of behavior pattern information from event information f) Judgment of unsteady state and steady state by comparison of event information and behavior pattern information g) Abnormality detection by detection logic and judgment of monitoring target person abnormality by judgment logic h) Generation of calling sound and sound output i) Processing for realizing functions such as generation of warning information is included, and one or more of these processes are executed.

<Hardware of watching unit 32 (= 32-1, 32-2, 32-3, 32-4)>
FIG. 6 shows the hardware of the watching unit 32. The monitoring unit 32 includes a voice signal input unit 40, a voice signal output unit 42, and a timer 44 in addition to the processing unit 10, the communication unit 12, the microphone 36, and the speaker 38 described above.
The audio signal input unit 40 amplifies the sound collection signal of the microphone 36 under the control of the processing unit 10 and inputs it to the processing unit 10. The audio signal output unit 42 reproduces the call audio signal generated by the processing unit 10 under the control of the processing unit 10 and outputs it to the speaker 38.
The timer 44 is used to measure the duration of the steady state, the duration of the abnormality of the watching target person, and the like under the control of the processing unit 10 using the counter. The timer 44 may be replaced with a soft timer that measures the time under the control of the processor 46.

The processing unit 10 includes a processor 46, a storage unit 48, and an input / output unit (I / O) 50. The processor 46 executes various processes such as an OS (Operating System) stored in the storage unit 48 and monitoring processing, such as sensor monitoring processing, abnormality monitoring / calling processing, and location determination processing, and executes the above-mentioned information processing. do. That is, the processor 46 includes functions as an audio processor, such as a state determination including digitization of an analog audio signal.

The storage unit 48 is an example of a recording medium for storing the program of the present disclosure, and includes a storage element such as a ROM (Read-Only Memory) and a RAM (Random-Access Memory). The storage unit 48 stores processing information for executing various processes such as an OS and monitoring processing, such as monitoring processing, abnormality monitoring / calling processing, and location determination processing. The storage unit 48 stores various information such as behavior information, whereabouts information, and voice information of the monitoring target person P. Monitoring management DB 52 (FIG. 13), behavior pattern information DB 74 (FIG. 14), event / behavior pattern comparison. Various DBs such as DB88 (FIG. 15) are constructed.
The I / O 50 takes in a sound collecting signal from the microphone 36 and inputs / outputs information such as an audio signal output to the speaker 38 under the control of the processor 46.

<Acquisition process of behavior pattern information>
FIG. 7 shows an example of a process of acquiring behavior pattern information. In this behavior pattern information acquisition process, life sound information acquisition (S201), acoustic analysis (S202), identification of the location L of the watching target person P (S203), event information acquisition (S204), and event information accumulation (S204). S205), acquisition of behavior pattern information of the watch target person P (S206), recording of behavior pattern information (S207), and the like are included.

Acquisition of living sound information (S201): Sounds are collected by the microphone 36 of the unit 32 while watching various sounds generated at the locations L1, L2, L3, and L4. That is, the location L1 is mainly the microphone 36 of the watching unit 32-1, the location L2 is mainly the microphone 36 of the watching unit 32-2, the location L3 is mainly the microphone 36 of the watching unit 32-3, and the location L4 is mainly the monitoring unit. Sound is collected by the microphone 36 of 32-4. In this case, sound is collected by the watching unit 32 adjacent to each location L, and the sound pressure level differs depending on the distance, but this sound pressure level information is also acquired as living sound information at the same time.

Acoustic analysis (S202): This acoustic analysis uses at least one or more of sound pressure, frequency, and period in the sound collection signal, and includes a process for detecting an event of the watch target person P, and this event. Is acquired as event information. In this case, it is specified from the acoustic characteristics of the acquired sound collection information whether or not the sound is emitted from the monitoring target person P. Then, when the level of the living sound information exceeds the threshold value, an abnormal state can be detected at this stage of acoustic analysis.
Identification of the location L of the monitoring target person P (S203): The location L of the monitoring target person P can be specified by the sound pressure level of the living sound information acquired from the location L. In this case, the location L may be specified by recording the comparison sound information of the living sound information and comparing the recorded information with the acquired information.

Acquisition of event information (S204): The processing unit 10 receives the sound collection signal of the microphone 36 and acquires the sound collection information, and acquires the sound collection information as event information by changing the level of the sound collection information.
Accumulation of event information (S205): The acquired event information is continuously recorded and accumulated in the monitoring management DB 52 (FIG. 13). This event information is accumulated in a predetermined period including, for example, a day, a week, a month, and a year.
Acquisition of behavior pattern information of the watching target person P (S206): The processing unit 10 acquires the behavior pattern information of the watching target person P from the accumulated event information group. The acquisition of this behavior pattern information includes a step of updating and optimizing the acquired behavior pattern information.
Recording of behavior pattern information (S207): This behavior pattern information is recorded in the behavior pattern information DB 74 (FIG. 14). This behavior pattern information is comparison information with the latest event information, and is used for determining an abnormality of the watching target person P.

<Coordination and priority of watching unit 32>
The monitoring units 32-1, 32-2, 32-3, and 32-4 simultaneously collect sounds individually at the locations L1, L2, L3, and L4, and acquire the behavior pattern information and the location information of the monitoring target person P. That is, all the information is shared by the watching units 32-1, 32-2, 32-3, and 32-4 in cooperation with each other.
When the location L of the watch target person P is specified, priority is given to the watch unit 32 at that position, the watch unit 32 performs the master operation, and the other watch units 32 shift to the slave operation. It may be realized.

<Acquisition process of behavior pattern information>
FIG. 8 is an example of acquisition processing of behavior pattern information of the person to be watched, FIG. 8A is a living sound information group, FIG. 8B is a specific information of a location L, FIG. 8C is an event information group, and FIG. D of 8 shows the behavior pattern information.
The living sound information group (A in FIG. 8) includes a plurality of living sound information A1, A2, ..., An. These living sound information A1, A2, ..., An are sound collecting information representing bathing collected in a specific period. Taking the living sound information A1 as an example, A11 is an operation sound for opening the door 30-1, A12 is an operation sound for closing the door 30-1, A13 is a mixed sound of a shower sound and a behavior sound of the watching target person P, and A14 is a mixed sound. The shower sound, A15, is an operation sound for opening the door 30-1.

The specific information of the whereabouts L (B in FIG. 8) represents the whereabouts L1 of the person to be watched over. In this case, the processing unit 10 recognizes from the mixed sound A13 including the operation sounds A11 and A12 of the door 30-1 and the shower sound that the watching target person P is the location L1 = bathroom 6-1.
The event information group (C in FIG. 8) includes a plurality of event information C1, C2, ..., Cn. The event information C1 is obtained from the living sound information A1, the event information C2 is obtained from the living sound information A2, ..., And the event information Cn is obtained from the living sound information An.
The behavior pattern information (D in FIG. 8) is a collection of a plurality of event information C1, C2, ..., Cn, and represents a behavior pattern at the location L1 of the watch target person P. This behavior pattern information is recorded in the behavior pattern information DB 74 (FIG. 14) and the event / behavior pattern comparison DB 88 (FIG. 15).

<Comparison process of behavior pattern information and latest event information>
FIG. 9 is an example of a process of comparing the behavior pattern information with the latest event information. In this comparison process, acquisition of living sound information (S301), acoustic analysis (S302), identification of the location L of the watch target person P (S303), acquisition of behavior pattern information (S304), acquisition of the latest event information (S305). ), Change determination (S306), timer 44 timekeeping addition (S307), timer reset (S308), duration determination (S309), comparison between the latest event information and behavior pattern information (S310), unsteady state (= Normal) determination (S311), steady state (= abnormal) determination (S312), and the like are included.

Acquisition of living sound information (S301): Upon receiving a sound collecting signal from the microphone 36 of each monitoring unit 32, the processing unit 10 continuously acquires living sound information from the monitoring area 4.
Acoustic analysis (S302): Since it is the same as S202 in FIG. 7, the description thereof will be omitted.
Identification of the location L of the monitoring target person P (S303): The processing unit 10 identifies the location L of the monitoring target person P from the acquired living sound information.
Acquisition of behavior pattern information (S304): Based on the identification of the watching target person P and the watching area 4 or the location L, the processing unit 10 acquires the behavior pattern information from the behavior pattern information DB 74.

Acquisition of the latest event information (S305): The processing unit 10 acquires the latest event information based on the acoustic analysis from the living sound information. That is, the event information of the watch target person P at the place L is acquired.
Determination of change (S306): The processing unit 10 acquires change information from living sound information and determines a change in acoustic information. This change information is, for example, information indicating a change such as whether the same level of sound is continuous or discontinuous.
Timekeeping addition of timer 44 (S307): The processing unit 10 controls the timekeeping of the timer 44 and measures the duration of the sound collection information without any change in sound.

Timer reset (S308): The processing unit 10 controls the timing of the timer 44, resets the duration if there is a change in the sound in the sound collection information, and returns to S302.
Determination of duration (S309): The processing unit 10 determines the duration accumulated in the timer 44. This duration is the time that continues from the last change of the sound in the sound collection information. In this determination of the duration, it is determined whether the duration is equal to or longer than a predetermined threshold time.
Comparison between the latest event information and the behavior pattern information (S310): The processing unit 10 compares the latest event information with the behavior pattern information acquired in S304, and the latest event information is different from the behavior pattern information. Judge whether or not.

Unsteady state (= normal) determination (S311): If the duration from the last change of sound is less than a predetermined threshold value (NO in S309), the processing unit 10 determines that it is unsteady state. If the latest event information matches the behavior pattern information (NO in S301), it is determined to be unsteady state. That is, there is a movement in the watching target person P from the sound collection information, and there is no abnormality in the watching target person.
Steady state (= abnormal) determination (S312): In the processing unit 10, the duration from the last change of sound is equal to or longer than a predetermined threshold time (YES in S309), and the latest event information is different from the behavior pattern information. If so (YES in S310), it is determined to be in a steady state. That is, there is no movement in the watching target person P from the sound collection information, and in this case, the watching target person may be abnormal or the watching target person may be absent.

<Anomaly detection logic processing example 1>
FIG. 10 shows an example 1 of this abnormality detection logic processing, which includes living sound information (A in FIG. 10), latest event information (B in FIG. 10), behavior pattern information (C in FIG. 10), and differences (match / mismatch). ) Judgment (D in FIG. 10) and judgment information (E in FIG. 10) are shown.
The living sound information (A in FIG. 10) is the living sound information acquired by the processing unit 10.
The latest event information (B in FIG. 10) is event information acquired from the living sound information (A in FIG. 10).
The behavior pattern information (C in FIG. 10) is the behavior pattern information acquired from the behavior pattern information DB 74 by the processing unit 10 based on the living sound information (A in FIG. 10) and the event information (B in FIG. 10). In this example, it is behavior pattern information (C in FIG. 10).

The difference (match / mismatch) determination (D in FIG. 10) indicates a determination of matching or mismatch between the latest event information (B in FIG. 10) and the behavior pattern information (C in FIG. 10). In the sections D1, D2, and D3, the latest event information matches the behavior pattern information (C in FIG. 10), whereas in the section D4, the latest event information does not match, indicating an abnormal state. That is, in the section D4, the event information (B in FIG. 10) is missing.
The judgment information (E in FIG. 10) represents this judgment result, and the judgment result representing the steady state (= abnormal) in the period when the event information and the behavior pattern information do not match, and the non-steady state (= normal) in the period when they match. can get. In other words, while taking a bath, there is an abnormality in the person being watched over.

<Anomaly detection logic processing example 2>
FIG. 11 shows an abnormality detection logic processing example 2, and shows life sound information (A in FIG. 11), event information (B in FIG. 11), and state determination (C in FIG. 11).
In this processing example 2, as shown in the living sound information (A in FIG. 11), the sound collecting information including the excessive sound information due to the overturning of the watching target person P is shown. If the person P to be watched falls over at the place L1 (= bathroom 6-1) or the like, a loud sound accompanied by vibration is generated.

The event information (B in FIG. 11) acquired from this living sound information includes information representing this event, in which an event is detected from an excessive sound generated by a fall of the watching target person P or the like.
In this case, as shown in A of FIG. 11, a threshold level Sth is set for the sound collection level, and if the sound level is equal to or higher than the threshold level Sth, an abnormal event is set and the determination state is determined to be a steady state (= abnormal). .. The threshold level Sth may be set to a level that should be determined to be abnormal with respect to the sound collection level.
The state determination (C in FIG. 11) is a determination result using the event information (B in FIG. 11). Therefore, in this state determination, a determination result indicating that the state has changed from the unsteady state to the steady state due to the occurrence of an event such as a fall has been obtained.

<Anomaly detection logic table 49>
FIG. 12 shows the abnormality detection logic table 49. The abnormality detection logic table 49 shown in FIG. 12 is an example, and the present disclosure is not limited to the contents of such a table.
The abnormality detection logic table 49 includes a pattern unit 51, an analysis information unit 53, and a state determination unit 55.
For example, "pattern A", "pattern B", "pattern C", ... Are stored in the pattern unit 51 as identification information of the detection pattern.

The analysis information unit 53 stores the analysis information which is the analysis result from the sound collection information. This analysis information includes sound information representing the analysis result such as "non-steady sound", "steady sound", and "loud sound above the threshold level". The “unsteady sound” is a sound that changes with time, and is a sound that represents the movement of the watching target person P. The "steady sound" is a sound representing a time-dependent state in which the sound does not change for X seconds or longer, for example, for the duration described above. That is, it is not the operating sound of the person P to be watched over, but the background sound that does not change, such as a continuous shower sound. Further, the "loud sound equal to or higher than the threshold level" is a sound representing an unexpected state such as a fall of the watching target person P.
The state determination unit 55 stores determination information representing the determination result by the processing unit 10. This determination result includes a non-steady state (= normal) and a steady state (= abnormal).

Therefore, the following abnormality detection results are presented in the abnormality detection logic table 49.
In the pattern A, the “non-steady sound” is obtained from the sound collection information, and the non-steady state, that is, the abnormality of the watching target person is not detected.
In pattern B, a "steady sound" is obtained from the sound collection information, and a steady state, that is, an abnormality of the watching target person is detected.
In pattern C, "a loud sound equal to or higher than the threshold level" is obtained from the sound collection information, and a steady state, that is, an abnormality of the watching target person is detected (= A in FIG. 11).

<Watching management DB52>
FIG. 13 shows an example of the monitoring management DB 52. The monitoring management DB 52 includes a management file 54 that stores monitoring information.
In this management file 54, the watching area information unit 56, the watching target person information unit 58, the date and time information unit 60, the sound collecting information unit 62, the whereabouts information unit 64, the event information unit 66, the judgment information unit 68, and the calling voice information unit 70 , History information unit 72 and the like are included.

The watching area information unit 56 stores attribute information representing the attributes of the watching area 4, morphological information representing the area of the house and the form of the place of residence, and the like.
The monitoring target person information unit 58 stores identification information such as the name of the watching target person P.
The date / time information unit 60 stores the date / time information when the abnormality detection is executed from the calendar information.
The sound collecting information unit 62 stores sound collecting information (= living sound information) representing the living sound acquired by the processing unit 10 from the sound collecting signal of the microphone 36.
The location information unit 64 stores location information indicating the location L, such as the moving location of the monitoring target person P in the monitoring area 4.

The event information unit 66 stores the event information acquired by the processing unit 10 from the living sound information by acoustic analysis.
The determination information unit 68 stores determination information such as a non-steady state, a steady state, and an abnormal sound pressure level, which are determination results by the processing unit 10. That is, determination information such as a non-steady state (= normal) and a steady state (abnormality) of the watch target person P is stored.
The call voice information unit 70 stores the call voice information generated from the processing unit 10 when the monitoring target person is abnormal. That is, notification information such as call voice information from the speaker 38 of the whereabouts L1, L2, L3, and L4 is stored.
The history information unit 72 stores history information representing the processing history, determination history, call voice generation and output, response history, and the like of the processing unit 10.

<Behavior pattern information DB74>
FIG. 14 shows an example of the behavior pattern information DB 74. The behavior pattern information DB 74 includes a management file 76 that stores the behavior pattern information.
The management file 76 includes a watching area information unit 78, a watching target person information unit 80, an accumulation period information unit 82, an event information unit 84, a behavior pattern information unit 86, and the like.

Similar to the management file 54 of the watching management DB 52, the watching area information unit 78 stores attribute information representing the attributes of the watching area 4, morphological information representing the area of the house and the form of the place L, and the like.
Similar to the management file 54 of the monitoring management DB 52, the monitoring target person information unit 80 stores identification information such as the name of the monitoring target person P.
The accumulation period information unit 82 stores period information indicating the accumulation period of event information.
The event information unit 84 stores the event information acquired by the processing unit 10 from the living sound information by acoustic analysis.
The behavior pattern information unit 86 stores the above-mentioned behavior pattern information calculated by the processing unit 10 from a plurality of event information within the accumulation period.

<Event / behavior pattern comparison DB88>
FIG. 15 shows an example of the event / behavior pattern comparison DB88. The event / behavior pattern comparison DB 88 includes a management file 90 that stores comparison information.
The management file 90 includes a monitoring area information unit 92, a monitoring target person information unit 94, a date / time information unit 96, an event information unit 98, a behavior pattern information unit 100, a comparison result information unit 102, a history information unit 104, and the like.

Similar to the management file 54 of the watching management D522, the watching area information unit 92 stores attribute information representing the attributes of the watching area 4, morphological information representing the area of the house and the form of the place L, and the like.
Similar to the management file 54 of the monitoring management DB 52, the monitoring target person information unit 94 stores identification information such as the name of the monitoring target person P.
The date and time information unit 96 stores date and time information indicating the monitoring date and time of the monitoring target person P.
The latest event information acquired from the living sound information is stored in the event information unit 98.
The behavior pattern information unit 100 stores behavior pattern information to be compared with the event information.
The comparison result information unit 102 stores result information representing the comparison result between the behavior pattern information and the latest event information.
Similar to the management file 54 of the monitoring management DB 52, the history information unit 104 stores history information representing the processing history, determination history, call voice generation and output, response history, and the like of the processing unit 10.

<Effect of one embodiment>
According to this embodiment, any of the following effects can be obtained.
(1) It is possible to perform simple and highly accurate behavior and abnormality detection of the watching target person P in each private space 6 by the living sound information of the watching area 4, and it is possible to construct a highly reliable watching system.
(2) A monitoring system can be easily realized without requiring a large-scale renovation of the house.
(3) The behavioral status of the watching target person P can be grasped from the living sound information, and abnormality detection can be performed without impairing the privacy of the watching target person P.
(4) The microphone 36 can be used to collect the sound in the location L with high sensitivity, and the sound collection information can be acquired to monitor the behavior of the target person P.
(5) Abnormality detection of the watching target person P is premised on the movement of the watching target person P and the presence of background sound. Since it is determined to be a state (abnormality), the abnormality of the watching target person P can be detected easily and with high accuracy.
(6) Since a call is made to the watch target person P, the watch target person P can respond to the call voice, and if there is no response, after a lapse of a predetermined time, an abnormality is detected from the outside by the call, and an unexpected situation occurs. Can be avoided.

(7) In the conventional human detection using a human detection sensor, etc., a malfunction or misjudgment occurred when a noise component was included in the element that triggers the judgment of normality / abnormality. Since the operation is detected by sound, it is possible to construct an excellent detection system that suppresses malfunctions.
(8) When monitoring with a camera, image processing becomes large in order to avoid impairing privacy, and there is a problem that it is difficult for the observer to make a judgment. However, such inconvenience can be avoided by the abnormality detection of the present disclosure. ..
(9) It is possible to realize the monitoring area 4 with the abnormality detection function of the present disclosure, and to provide a safe and secure living space.
(10) The watching area 4 disclosed in the above embodiment can be set to the living space L of the private space 6 such as the living space of the condominium in addition to the detached house to watch over the watching target person P and detect an abnormality.

<Variations included in one embodiment and their effects>
This embodiment includes the following variations.
(1) A control may be added in which the response voice from the watching target person P who has received the calling voice is collected by the microphone 36, and the transition from the steady state to the non-steady state is performed using the sound collection information.
(2) The monitoring of the behavior information of the person P to be watched may include any one of the loudness, sound pressure, frequency, and period of the living sound, or two or more of them.
(3) A system may be constructed in which a questioning voice is output from an information terminal notified by the watching system 2.
(4) Regarding the watching units 32-1, 32-2, 32-3, and 32-4, the server function is set in the watching unit 32-1 to manage the watching units 32-2, 32-3, and 32-4. Alternatively, a server computer may be provided to supervise the monitoring units 32-1, 32-2, 32-3, and 32-4.

<Abnormality report to the outside by the communication unit 106>
For example, as shown in FIG. 16, the communication unit 106 may be provided in the above embodiment.
When the processing unit 10 determines an abnormality of the person to be watched over, the determination information can be reported from the communication unit 106 using a communication medium such as a public line. For example, information such as a smartphone carried by the administrator. It is possible to notify the terminal 108 that the monitoring target person P is abnormal or has made a call.

<Behavior detection by door sensors 107-1, 107-2, 107-3, 107-4>
In the above embodiment, the microphone 36 is used to detect the behavior of the location L and the watching target person P. However, as shown in FIG. 17, the door sensors 107-1, 107-2, 107 that detect the opening / closing of the door 30 are detected. -3, 107-4 may be provided.
The entry, exit, and presence of the room are monitored by the detection outputs of the door sensors 107-1, 107-2, 107-3, and 107-4, and the monitoring units 32-1, 32-2, at each of the locations L1, L2, L3, and L4. The sound collection information acquired from the sound collection signals of the microphones 36 of 32-3 and 32-4 may be combined to monitor the location L and the presence / absence of the monitoring target person P.

<Watching process 109>
FIG. 18 shows the monitoring process 109 according to the embodiment of the present disclosure. This embodiment is an example of a program for realizing the monitoring of the present disclosure, and the program of the present disclosure is not limited to such an embodiment.
The watching process 109 includes a watching unit management process 110, an abnormality monitoring / calling process 112, and a whereabouts determination process 114.
The monitoring unit management process 110 includes an operation confirmation function 116 of the monitoring unit 32, an activation status confirmation function 118 of the abnormality monitoring / calling process 112, an activation instruction function 120 of the abnormality monitoring / calling process 112, and an activation instruction function 122 of the location determination process 114. And so on, it starts every predetermined time, for example, every second.

The abnormality monitoring / calling process 112 includes an abnormality detection logic function 124, an abnormality confirmation function 126, a response confirmation function 128, a call by the speaker 38, a time measurement and an elapsed time determination function 130, and the like.
The whereabouts determination process 114 includes the whereabouts determination logic function 132, the whereabouts confirmation function 134, the end instruction function 136 of the whereabouts determination process 114, the end instruction function 138 of the abnormality monitoring / calling process 112, and the like.

<Watching unit management process 110>
FIG. 19 shows an example of a processing procedure by the monitoring unit management process 110.
The watching unit management process 110 is started every second, for example, and confirms the operating state of the watching unit 32 (= 32-1, 32-2, 32-3, 32-4).
If the monitoring unit 32 is in the stopped state (while S401 is stopped), the abnormality monitoring / calling process 112 and the whereabouts determination process 114 are terminated (S402).
If the monitoring unit 32 is in the operating state (during the operation of S401), the activation state of the abnormality monitoring / calling process 112 is confirmed (S403).
If the abnormality monitoring / calling process 112 is not in the activated state (S403 is not started), the abnormality monitoring / calling process 112 and the whereabouts determination process 114 are started (S404).
If the abnormality monitoring / calling process 112 is in the activated state (S403 has already been activated), the monitoring unit management process 110 is terminated (S405), and this process is completed.

<Abnormality monitoring / calling process 112>
FIG. 20 shows an example of the processing procedure of the abnormality monitoring / calling process 112.
When this abnormality monitoring / calling process 112 is activated, the abnormality detection logic is executed (S501), and it is determined whether or not the monitoring target person P has an abnormality (S502). If there is no abnormality in the watch target person P (no abnormality in S502), the execution of the abnormality detection logic is continued.
If there is an abnormality in the watch target person P (there is an abnormality in S502), the timer 44 is timed together with the call from the speaker 38 (S503).

It is determined whether or not there is a response from the watching target person P to this call (S504). If there is a response (YES in S504), the process returns to S501. The presence or absence of this response may be determined by using the operation sound in addition to the utterance from the watching target person P.
If there is no response (NO in S504), it is determined whether the measurement time of the timer 44 has elapsed the predetermined time (S505). When the predetermined measurement time has elapsed (YES in S505), the timer 44 is timed together with the call from the speaker 38 (S506).
It is determined whether the measurement time of the timer 44 has elapsed (S507), and if the predetermined time has elapsed (YES in S507), the process returns to S506 and the timer 44 is clocked again with the call from the speaker 38 (S506). .. Therefore, if the abnormal state continues, the speaker 38 continuously calls at predetermined time intervals.

<Location judgment process 114>
FIG. 21 shows an example of the processing procedure of the location determination process 114.
When this location determination process 114 is activated, the location determination logic is executed (S601), and it is determined whether or not the user is in the location L (= any of L1, L2, L3, and L4) (S602).
If there is a person P to be watched over (YES in S602), the whereabouts determination process 114 is terminated (S603).
If there is no monitoring target person P (NO in S602), the monitoring unit management process 110 and the abnormality monitoring / calling process 112 are terminated (S604).

<Hardware of watching unit 32 (= 32-1, 32-2, 32-3, 32-4)>
FIG. 22 shows an example of the hardware of the monitoring unit 32 according to the embodiment of the present disclosure. In FIG. 22, the same parts as those in FIG. 5 are designated by the same reference numerals, and the description thereof will be omitted.
The monitoring unit 32 of this embodiment is provided with a communication unit 212 together with a circuit unit 210 such as a Raspberry Pi, for example. The circuit unit 210 constitutes a circuit including a processing unit 10, a timer 44, an audio signal input unit 40, an audio signal output unit 42, a power supply unit, and the like. The communication unit 212 constitutes the communication units 12 and 106.

The circuit unit 210 is provided with USB ports 214-1, 214-2, 214-3, 214-4, 214-5, connection ports 214-6, and power supply ports 214-7.
A speaker 38 is connected to the USB port 214-1 and a microphone 36 is connected to the USB port 214-2. As a result, the sound collection signal is taken in and the calling voice signal is output.
A storage card 216 is detachably attached to the USB port 214-5. The storage card 216 is an example of a recording medium, and is an example of the storage unit 48 described above.
The door sensor 107 (= 107-1, 107-2, 107-3, 107-4) described above may be connected to the connection port 214-6 to capture the sensor output.
A power cord 218 is connected to the power port 214-7, and a power connector 220 at the end receives power from, for example, a commercial AC power supply.

<Watching unit 32>
FIG. 23 shows an example of the watching unit 32 (= 32-1, 32-2, 32-3, 32-4). In FIG. 23, the same parts as those in FIG. 22 are designated by the same reference numerals.
Watching units 32 (= 32-1, 32-2, 32-3, 32-4) are attached to the wall portion 224 adjacent to the upper corner portion of the door frame 222. The monitoring unit 32 includes a water resistant material, for example, a fixing base 226 made of a synthetic resin and a cover portion 228. The fixed base 226 is fixed to the wall portion 224 that partitions the place L, and the cover portion 228 is removable from the fixed base 226.

A sound collecting port 232 and a sound emitting port 234 are formed in the cover portion 228, sound is collected from the sound collecting port 232 to the microphone 36, and the call sound of the speaker 38 is output from the sound collecting port 234. The power connector 220 connected to the power cord 218 pulled out from the cover portion 228 is attached to the adjacent power outlet 236.

<Effect of one example>
According to this one embodiment, any of the following effects can be obtained.
(1) In this embodiment, the abnormality detection method according to one embodiment can be realized, and the monitoring system 2 can be constructed.
(2) It can be installed incidentally in an existing house, etc. to watch over the person P to be watched and detect its abnormality.
(3) It can be realized by simple construction without carrying out major remodeling work on the house.
(4) It is possible to realize and provide a house equipped with the watching system 2.

[Other Embodiments]
This disclosure includes any of the following:
(1) This disclosure applies not only to general homes but also to long-term care facilities.
(2) In the above embodiment, the microphone 36 and the speaker 38 are separate units, but a single unit may have both a sound collecting function and a sound emitting function.
(3) Since the present disclosure uses the microphone 36 to detect the whereabouts of the person P to be watched over, a sounding device that produces different sound effects for each door 30-1, 30-2, 30-3, and 30-4 is provided. The opening and closing of the doors 30-1, 30-2, 30-3, and 30-4 may be detectable.
(4) A motion sensor or a door sensor 107 may be detachably connected to the monitoring unit 32.
(5) A door sensor information unit is provided in the management file 54 of the monitoring management DB 52, and only the opening / closing information, the entry information, the exit information, and the opening / closing detection indicating the opening / closing of the door sensors 107-1, 107-2, 107-3, and 107-4 are used. The occupancy information and the like may be stored and used for the whereabouts information of the person P to be watched over.

As described above, the most preferable embodiments of the present disclosure have been described. The present disclosure is not limited to the above description. Various modifications and modifications can be made by those skilled in the art based on the gist of the invention described in the claims or disclosed in the form for carrying out the invention. It goes without saying that such modifications and changes are included in the scope of the present disclosure.

According to the present disclosure, it is possible to monitor the whereabouts of the person being watched over in a private space such as a house, its bathroom, a dressing room, a toilet, and a bedroom. Since it is registered as information, it is possible to easily and highly accurately detect an abnormality that occurs in the person being watched over, and it is useful because it is possible to speed up the detection of an abnormality in the person being watched over, a call to the person being watched over, and an abnormality report.

2 Watching system 4 Watching area 6 Private space 6-1 Bathroom 6-2 Dressing room 6-3 Toilet 6-4 Bedroom 8 Sound collecting section 10 Processing section 12 Communication section 14 Information input section 16 Information presentation section 18 Watching area Information acquisition section 20 Watching target person information acquisition unit 22 Living sound information acquisition unit 24 Event information acquisition unit 25 Behavior pattern information acquisition unit 26 Notification unit 28 Abnormality judgment unit 30-1, 30-2, 30-3, 30-4 Doors 32, 32 -1, 32-2, 32-3, 32-4 Watching unit 34 Network 36 Microphone 38 Speaker 40 Audio signal input 42 Audio signal output 44 Timer 46 Processor 48 Storage 50 Input / output (I / O)
52 Watching management DB
54 Management file 56 Watching area information department 58 Watching target person information department 60 Date and time information department 62 Sound collecting information department 64 Whereabouts information department 66 Event information department 68 Judgment information department 70 Calling voice information department 72 History information department 74 Behavior pattern information DB
76 Management file 78 Watching area information section 80 Watching target person information section 82 Accumulation period information section 84 Event information section 86 Behavior pattern information section 88 Event / behavior pattern comparison DB
90 Management file 92 Watching area information section 94 Watching target person information section 96 Date and time information section 98 Event information section 100 Behavior pattern information section 102 Comparison result information section 104 History information section 106 Communication section 107-1, 107-2, 107-3 , 107-4 Door sensor 108 Information terminal 109 Watching processing 110 Watching unit management processing 112 Abnormality monitoring / calling processing 114 Location judgment processing 116 Operating status confirmation function 118 Startup status confirmation function 120 Startup instruction function 122 Startup instruction function 124 Abnormality detection logic Function 126 Abnormality confirmation function 128 Response confirmation function 130 Time measurement and elapsed time judgment function 132 Location judgment logic function 134 Location confirmation function 136 End instruction function 138 End instruction function 210 Circuit unit 212 Communication unit 214-1, 214-2, 214 -3, 214-4, 214-5 USB port 214-6 Connection port 214-7 Power port 216 Storage card 218 Power cord 220 Power connector 222 Door frame 224 Wall 226 Fixing base 228 Cover 230 Wiring 232 Sound collecting port 234 Sound outlet 236 Power outlet

Claims (22)

A process in which a private space including one of a bathroom, a dressing room, a toilet, a bedroom, or two or more of them is used as a watching area, and the watching area information acquisition unit acquires watching area information representing the watching area.
The process by which the watching target person information acquisition department acquires the watching target person information representing the watching target person, and
The process in which the living sound information acquisition department collects the living sounds of the person being watched over in the private space and acquires the living sound information.
A process in which the event information acquisition unit detects the behavior of the watching target person from the living sound information and acquires event information representing the behavior.
The behavior pattern information acquisition unit collects the event information to generate the behavior pattern of the watching target person, and acquires the behavior pattern information representing the behavior pattern related to the watching area information and the watching target person information. And the process to do
The abnormality determination unit compares the behavior pattern information with the latest event information, monitors the match / mismatch between the two, and monitors the duration of the mismatch state. If at least the duration is equal to or longer than the threshold time, The process of determining that the person to be watched is abnormal,
A watching method characterized by including. Further, if the abnormality determination unit includes event information in which the latest event information matches the event information existing in the behavior pattern information, it is in a non-steady state, and if there is no event information, it is in a steady state, and the duration of the steady state is set. The monitoring method according to claim 1, further comprising a step of determining the abnormality of the monitoring target person when the value exceeds the threshold time or includes excessive sound information that does not exist in the behavior pattern information. Further, the behavior pattern information unit accumulates the event information for a predetermined period including a day, a week, a month, and a year to generate the behavior pattern information, or updates the event information to generate the behavior pattern information. The monitoring method according to claim 1 or 2, wherein the monitoring method includes a step of performing. Further, when the abnormality determination unit determines the abnormality of the monitoring target person, the reporting unit calls to the location of the monitoring target person and outputs voice information.
When there is no response from the watching target person and the duration from the output of the calling voice exceeds the threshold time, the reporting unit outputs alarm information, and
The monitoring method according to any one of claims 1 to 3, wherein the monitoring method comprises. Further, claims 1 to 2, wherein the event information acquisition unit includes a step of acquiring the event information by using at least one or two or more of the sound pressure, frequency, and period of the living sound information. The monitoring method according to any one of claims 4.
A watching area information acquisition unit that acquires watching area information representing the watching area, which is a private space including one of the bathroom, dressing room, toilet, and bedroom, or two or more of them.
The watching target person information acquisition department that acquires the watching target person information representing the watching target person,
The life sound information acquisition department that collects the life sounds of the person being watched over in the private space and acquires the life sound information,
An event information acquisition unit that detects the behavior of the watching target person from the living sound information and acquires event information representing the behavior.
A behavior pattern information acquisition unit that collects the event information to generate a behavior pattern of the watching target person, and acquires the behavior pattern information representing the behavior pattern associated with the watching area information and the watching target person information. ,
By comparing the behavior pattern information with the latest event information, the match / mismatch between the two is monitored, and the duration of the mismatch state is monitored. Abnormality judgment unit and
A watching system characterized by being equipped with. Further, the abnormality determination unit sets a non-steady state if the latest event information includes event information that matches the event information existing in the behavior pattern information, and a steady state if there is no event information, and the duration of the steady state. The monitoring system according to claim 6, wherein when the number exceeds the threshold time, or when the behavior pattern information includes excessive sound information that does not exist, the monitoring target person abnormality is determined. Further, the claim is characterized in that the behavior pattern information unit accumulates the event information in a predetermined period including a day, a week, a month, and a year, or updates the event information to generate the behavior pattern information. The monitoring system according to claim 6 or 7. Further, when the abnormality determination unit determines the abnormality, the call voice information is generated at the location of the watch target person, and there is no response from the watch target person, and the duration from the output of the call voice information is the reference. The monitoring system according to any one of claims 6 to 8, wherein when the time is exceeded, alarm information indicating the abnormality is generated. Further, claim 6 to 9, wherein the behavior pattern information unit acquires the event information from the living sound information using at least one or more of sound pressure, frequency, and period. The monitoring system described in any of the claims.
It ’s a program that you want your computer to run.
A function to obtain watching area information representing the watching area as a watching area in a private space including any one of a bathroom, a dressing room, a toilet, and a bedroom, or two or more of them.
A function to acquire information on the person to be watched, which represents the person to be watched, and
A function to collect the living sounds of the person being watched over in the private space and acquire the living sound information,
A function of detecting the behavior of the watching target person from the living sound information and acquiring event information representing the behavior, and
A function of collecting the event information to generate a behavior pattern of the watching target person, and acquiring the watching area information and the behavior pattern information representing the behavior pattern related to the watching target person information.
By comparing the behavior pattern information with the latest event information, the match / mismatch between the two is monitored, and the duration of the mismatch state is monitored. Judgment function and
A program for causing the computer to execute. Further, if the latest event information includes event information that matches the event information existing in the behavior pattern information, it is in a non-steady state, and if there is no event information, it is in a steady state, and the duration of the steady state exceeds the threshold time. The program according to claim 11, wherein the computer is made to execute the function of determining the abnormality of the watching target person when the behavior pattern information includes excessive sound information that does not exist. Further, the computer executes a function of accumulating the event information for a predetermined period including a day, a week, a month, and a year to generate the behavior pattern information, or updating the event information to generate the behavior pattern information. The program according to claim 11 or 12. Further, when the abnormality of the watching target person is determined, the reporting unit outputs the call voice information to the location of the watching target person, and the duration from the output of the calling voice without a response from the watching target person. The program according to any one of claims 11 to 13, for causing the computer to execute a function of outputting alarm information by the reporting unit when the threshold time is exceeded. Further, any one of claims 11 to 14 for causing the computer to execute a function of acquiring the event information by using at least one or two or more of the sound pressure, frequency, and period of the living sound information. The program described in the section. A recording medium comprising storing the program according to any one of claims 11 to 15. A watching device that serves as a watching area for a private space including one of a bathroom, a dressing room, a toilet, and a bedroom, or two or more of them.
The device unit is provided with a watching area information acquisition unit that acquires watching area information representing the watching area, and the device unit.
The watching target person information acquisition department that acquires the watching target person information representing the watching target person,
The life sound information acquisition department that collects the life sounds of the person being watched over in the private space and acquires the life sound information,
An event information acquisition unit that detects the behavior of the watching target person from the living sound information and acquires event information representing the behavior.
A behavior pattern information acquisition unit that collects the event information to generate a behavior pattern of the watching target person, and acquires the behavior pattern information representing the behavior pattern associated with the watching area information and the watching target person information. ,
By comparing the behavior pattern information with the latest event information, the match / mismatch between the two is monitored, and the duration of the mismatch state is monitored. Abnormality judgment unit and
A watching device that is characterized by being equipped with. Further, the abnormality determination unit sets a non-steady state if the latest event information includes event information that matches the event information existing in the behavior pattern information, and a steady state if there is no event information, and the duration of the steady state. The monitoring device according to claim 17, wherein when the number exceeds the threshold time, or when the behavior pattern information includes excessive sound information that does not exist, the monitoring target person abnormality is determined. Further, the claim is characterized in that the behavior pattern information unit accumulates the event information in a predetermined period including a day, a week, a month, and a year, or updates the event information to generate the behavior pattern information. The monitoring device according to claim 17 or claim 18. Further, when the abnormality determination unit determines the abnormality, the call voice information is generated at the location of the watch target person, and there is no response from the watch target person, and the duration from the output of the call voice is the reference time. The monitoring device according to any one of claims 17 to 19, wherein when the above value is exceeded, alarm information indicating the abnormality is generated. Further, claim 17 to 20, wherein the event information acquisition unit acquires the event information from the living sound information using at least one or more of sound pressure, frequency, and period. Watching device according to any claim. The monitoring method according to any one of claims 1 to 5, the monitoring system according to any one of claims 6 to 10, and the claim according to any one of claims 11 to 15. A residence comprising the program according to claim 16, the recording medium according to claim 16, and the monitoring device according to any one of claims 17 to 21.

Images