JP2003099878A - Human body abnormality detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To Provide a human body abnormality detector capable of surely monitoring a human body by retaining a detecting precision even if environmental conditions in a detection space is inferior. SOLUTION: This human abnormality detector is provided with a microphone 3a disposed in a bathroom 2 and detecting a sound issued in the bathroom 2, an infrared ray distance surveying sensor 3b detecting the distance to a testee, and first and second abnormality determination means 140 and 170 determining the abnormal action state of the testee based on a sound waveform expressing the sound detected by the microphone 3a and the moving distance of the testee found from the distance detected by the infrared distance surveying sensor 3b.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a human body abnormality detecting device for detecting an abnormal operation state of a person to be detected in a detection space.

[0002]

2. Description of the Related Art Conventionally, as this type of human body abnormality detection device, there is known a device which monitors an image in a detection space outside the space and detects that the detected person in the image has stopped moving and issues an alarm. Has been. In this conventional device, a camera is generally arranged at a position where the entire target space can be viewed, and the motion of the detected person is monitored while displaying an image captured by the camera on a monitor receiver. Then, an alarm is issued by detecting that the subject captured in the image has stopped moving.

In recent years, there has been an increasing social demand for caring for the elderly and sick and ill, and as one of the support systems therefor, the movement of the detected person in the living environment, especially in the private room space such as a bathroom or toilet, is monitored. However, there is a demand for a human body abnormality detection device for notifying a person in another space when the detected person suddenly stops moving due to an accident such as cardiac arrest.

[0004]

However, in order to apply it to those private room spaces, in a device in which a camera is installed to monitor with an image, it is applied to a dark room or a bathroom where steam is generated as a detection space. in case of,
There is a problem that the detection accuracy deteriorates.

Therefore, in view of the above points, an object of the present invention is to
Provided is a human body abnormality detection device capable of maintaining detection accuracy and reliably monitoring even when environmental conditions in a detection space are bad.

[0006]

In order to achieve the above object, the technical means described in claims 1 to 16 is adopted. That is, in the invention described in claim 1, a detection means (3a) for detecting sound or vibration in the detection space, and an entrance detection means (3b) for detecting the entrance of the person into the detection space. The room entrance detecting means (3b) detects the person to be detected and starts a monitoring operation, and based on the first information indicating the sound or vibration detected by the detecting means (3a), the abnormal operation state of the person to be detected is detected. First determining means (140,
And 140a).

According to the first aspect of the invention, based on the first information indicating the sound or vibration generated by the detected person,
First determination means (14) for determining an abnormal operation state of the detected person
0, 140a), it is possible, for example, to detect a sound that is louder in a closed space or in a narrow space and intrudes from outside the detection space as compared with a conventional monitoring device using a camera. Can easily detect the presence or absence of sound because it is easily attenuated by propagation in the air and transmission through wall surfaces. Especially when it is applied to a dark room or a bathroom where steam is generated, the detection accuracy is superior to the image using a camera.

In a camera, the narrower the space, the more restrictive the image range to be detected. Therefore, the mounting position must be restricted, or the camera must have an additional function such as tracking means. There is no restriction on the mounting position and there is no need to follow the directivity.

The invention according to claim 2 is characterized in that the entrance detecting means (3b) is a distance measuring means (3b) for detecting the distance to the detected person.

According to the second aspect of the present invention, by using the distance measuring means (3b) for detecting the distance to the detected person, it is possible to move the detected person into and out of the detection space up to the detected person. Since it is detected as a change in distance, that is, movement of the detected person, it is possible to reliably detect the entering and leaving of the detected person.

In the third and fourth aspects of the invention, the first determining means (140, 140a) is such that the first information indicating sound or vibration is in a substantially silent state or a non-vibrating state for a predetermined time. Is detected to determine the abnormal operation state of the detected person, specifically, the second information indicating the movement of the detected person is obtained from the distance information detected by the distance measuring means (3b). Second determination means (170) for determining the abnormal operation state of the detected person based on the second information is provided.
The judging means (170) detects that there is substantially no change based on the second information, and detects that the first information indicating sound or vibration is in a substantially silent state or a non-vibrating state for a predetermined time. The feature is that the abnormal operation state of the person who detects the object is determined.

According to the third and fourth aspects of the invention, the second information indicating the movement of the detected person does not substantially change,
In addition, by detecting that the sound of the first information is in a substantially silent state or a non-vibrating state for a predetermined time, in the second information, for example, the movement of the person to be detected has stopped and the body has stopped moving. Then, the abnormal operation state can be more surely determined by detecting the sound that is substantially silent.

In the fifth aspect of the invention, the first determination means (140, 140a) makes a sound detected by the detection means (3a) when the entrance detection means (3b) does not detect the person to be detected. Alternatively, the background noise may be provided as a reference value for detecting that the first information indicating the vibration is stored as background noise information or background vibration information as the background noise information or background vibration information. It is characterized by using information or dark vibration information.

According to the fifth aspect of the invention, specifically, the storage means (120) for storing the first information being monitored as background noise information or background vibration information when the detected person is not detected. ) Is used as a reference value for comparison, whereby it is easy to determine the abnormal operation state.

According to the sixth aspect of the invention, the first judging means (140, 140a) is configured to detect the sound or vibration detected by the detecting means (3a) based on the first information, and the first filter means (5a). ), A specific frequency component is extracted, and it is detected that it is in a substantially silent state or a non-vibration state, and the abnormal operation state of the subject is determined.

According to the invention described in claim 6, the specific frequency component is extracted from the first information by the first filter means (5a), and it is detected that there is substantially no sound or no vibration, and the detected object is detected. By determining the abnormal operation state of the person, the specific frequency component is determined by, for example, a bandpass filter from the first information of the sound or vibration generated in the detection space without being affected by the sound or vibration entering from the outside of the detection space. By extracting, the presence or absence of sound or vibration in the detection space can be detected more accurately and easily.

In a seventh aspect of the present invention, the first filter means (5a) is configured to switch at least two or more components as the specific frequency components to be extracted, or simultaneously extract. Is characterized by.

According to the seventh aspect of the present invention, at least two or more specific frequency components are switched and extracted from the sound or vibration generated in the detection space, or simultaneously extracted, so that the detection space is detected. Even if the resonance or the resonance frequency changes due to the change in the size, the sound or the vibration generated in the detection space can be reliably detected.

In the invention described in claim 8, the first judging means (140, 140a) is the first filter means (5a).
Of the specific frequency components extracted by the above, a sound or vibration continued for a predetermined threshold value (L1) or more and a third predetermined time (T3) or more is used as a detection sound or a detection vibration and is not in a substantially silent state or a non-vibration state. Is detected to determine the abnormal operation state of the detected person.

According to the eighth aspect of the present invention, among sounds or vibrations generated in the detection space, only sounds or vibrations that are relatively large and have a long duration are extracted to intrude from outside the detection space. Even if the specific frequency component extracted by the first filter means (5a) is not attenuated and remains in the sound or vibration generated, the sound or vibration in the detection space can be accurately and easily detected.

In the ninth aspect of the present invention, the first determining means (140, 140a) detects a sound or a vibration continued for a predetermined first threshold value (L1) or more and a third predetermined time (T3) or more. Even if it is detected as a sound or a detected vibration, if the detected sound or the detected vibration is not detected again within a predetermined range of time (T4 to T5) thereafter, it is detected that it is in a substantially silent state or a non-vibrated state. Is characterized by determining an abnormal operation state of the detected person.

According to the invention described in claim 9, for example, in a dressing room or a washroom adjacent to a bathroom, a single sound or a single vibration that vigorously closes a door or a door of a closet must be reproduced thereafter. For example, to determine that it is in a virtually silent state or in a vibrationless state, even if it detects a single sound or vibration from outside the detection space described above, it is determined to be in a substantially silent state or a vibrationless state. Erroneous detection can be prevented when determining the abnormal operation state of the detected person.

According to the tenth aspect of the present invention, the detection space is a bathroom, and hot water supply equipment for supplying hot water in the bathroom and hot water supply operation information indicating operation and stop of the hot water supply equipment are provided.
When the hot water supply operation information indicating the operation of the hot water supply device is input, the first determination means (140, 140a) causes the predetermined first
The threshold value (L1) or / and the third predetermined time period (T3) is changed to the second threshold value (L2) or / and the sixth predetermined time period (T6), and it is detected that there is substantially no sound or no vibration, The feature is that the abnormal operation state of the detected person is determined.

According to the tenth aspect of the present invention, when the hot water supply operation information indicating the operation of the hot water supply apparatus is input, specifically, the person to be detected puts hot water in the bathtub or the washing room. When a hot water supply operation is performed by opening a faucet or a shower such as when taking a shower at, the sound or vibration in the detection space is different, so the second threshold (L2) and / or the sixth predetermined time (T
By changing to 6), even when the faucet or shower is still open, it is superimposed on repetitive sound or repetitive vibration without continuous motion of the subject. Therefore, it is possible to reliably determine the abnormal operation state of the detected person by detecting only the operation sound of the detected person and the vibration caused by the operation.

In the invention described in claim 11, the first judging means (140, 140a) is the first filter means (5).
The specific frequency component extracted in a) is AC / DC converted with a predetermined first time constant into a direct current component, and AC / DC converted with a second time constant longer than the first time constant, and a variation threshold value. Then, a fixed threshold is added to this, and when the direct current component exceeds the combined threshold for a seventh predetermined time (T7) or more, it is determined that it is not in a silent state or a vibrationless state. .

According to the eleventh aspect of the present invention, the specific frequency component extracted by the first filter means (5a) is first AC / DC converted with the first time constant,
A suddenly fluctuating signal is used as a relaxed DC component, and a sound or vibration for a short time or less fluctuation is removed.

At the same time, by performing the AC / DC conversion with the second time constant longer than the first time constant, the first time constant A
A waveform that changes slightly more slowly than the C / DC-converted waveform is generated and used as a fluctuation threshold value, and a fixed threshold value is added to this to obtain a combined threshold value.

When the direct current component exceeds the synthesis threshold value for the seventh predetermined time (T7) or more, it is determined that the silent state or the non-vibrating state is not present, so that the silent state or the non-vibrating state continues for a certain time or more. After that, a relatively small sound or vibration can be detected, but a sound or vibration with little change, such as when the shower or faucet is left open, is detected only immediately after the occurrence and not thereafter. You can

In the twelfth aspect of the invention, the notifying means (150, 15) for notifying the subject when the first determining means (140, 140a) determines the abnormal operating state of the subject.
0a), and the first determination means (140, 140a)
After notifying the detected person by the notification means (150, 150a) when the abnormal operation state of the detected person is determined,
Response determination means (16) for determining whether or not the detected person has an abnormality
It is characterized by having 0).

According to the twelfth aspect of the invention, the mutual confirmation with the detected person is provided by including the response judgment means (160) for judging the presence or absence of abnormality of the detected person after notifying the detected person. It becomes possible to perform a stepwise human body abnormality detection device that is capable of performing the above-mentioned operation and has a dialogue with the person to be detected.

According to the thirteenth aspect of the present invention, the sound output from the notification means (150, 150a) is the sound from which the specific frequency component extracted by the first filter means (5a) has been removed in advance. It has a feature.

According to the thirteenth aspect of the invention, the voice notified in the detection space is the first filter means (5a).
Since the specific frequency component extracted by is previously removed or attenuated and notified, only the operation sound of the subject can be reliably detected by the detection means (3a) even during the notification.

In the fourteenth aspect of the present invention, when the first determination means (140, 140a) determines the abnormal operation state of the person being detected in the detection space, the alarm means (outputs an alarm outside the bathroom. 180, 200).

According to the fourteenth aspect of the present invention, the detection space is provided, for example, in a bathroom, so that there is no fear of causing sexual dissatisfaction in comparison with an image using a conventional camera. In addition, it is possible to provide a device that can detect an unexpected accident such as cardiac arrest while taking a bath.

In the fifteenth aspect of the invention, the notifying means (150, 150a) and the alarm means (180, 20).
0) is a loudspeaker (4b) or an alarm device (3) that outputs a voice.
c), the detection means (3a) is composed of a microphone (3a), and a loudspeaker (4b).
A second microphone (4d) for inputting a voice is added in the vicinity of, so that a communication means capable of bidirectional communication is provided inside and outside the detection space.

According to the fifteenth aspect of the invention, specifically, the second microphone (4d) is constituted by the loudspeaker (4b) for inputting or outputting voice, the alarm (3c) and the microphone (3a). By adding
By providing a two-way call means,
A low-cost call means can be configured, and conversations can be made in both directions, so that the detected person can be alerted before an abnormality occurs.

According to the sixteenth aspect of the present invention, of the voice input by the microphone (3a), the voice from which the specific frequency component is removed by the second filter means (5b) is output from the loudspeaker (4b). Is characterized by.

According to the sixteenth aspect of the present invention, it is possible to remove the components amplified by the reverberation sound and the resonance sound in the detection space from the sound emitted in the detection space. Therefore, for example, a shower or a washbasin can be used. Even if you make a big noise like throwing away hot water stored in a container, there is no sound crackling or annoying sound, and you can monitor accurately outside the detection space.

The reference numerals in parentheses of the above-mentioned means indicate the correspondence with the concrete means of the embodiments described later.

[0040]

BEST MODE FOR CARRYING OUT THE INVENTION (First Embodiment) A first embodiment of the human body abnormality detecting device of the present invention will be described below with reference to FIGS. First, FIG. 1 is an explanatory diagram of mounting a human body abnormality detecting device 1 in which the present invention is applied to a bathroom as a detection space.

The human body abnormality detecting device 1 is installed in a remote control panel 3 installed on a wall surface in a bathroom 2 which is a detection space, and in a living room outside the bathroom 2 where a person such as a living room or a kitchen (not shown) is always present. It is composed of an operation panel 4. In the remote control panel 3, a microphone 3a, which is a detection means for monitoring the sound in the bathroom 2, and an infrared detection sensor, which is a distance measurement means for detecting the distance to the person to be detected as an example of the entrance detection means 3b and an alarm device 3c for notifying the subject by a generated signal such as a voice.

The infrared distance measuring sensor 3b detects whether the person to be detected enters or leaves the bathroom 2, and the distance from the person to be detected allows the person to detect the inside of the bathroom 2 in the washing room 2b or the bath 2c. It is a sensor for detecting movement in
As shown in, the bathroom door 2a, the washroom 2b, and the bathtub 2c are preferably attached to the wall surface arranged in this order. Further, when the wall surface is other than the wall surface, it is preferable that the irradiation direction is freely adjusted toward the bathroom door 2a, and that the space between the washing room 2b and the bathtub 2c is required to be located.

In the present embodiment, the alarm device 3c is a notification means for notifying the person to be detected by a voice signal such as "Is it okay?"

Next, on the operation panel 4, an operation switch 4a for operating the human body abnormality detecting device and a loudspeaker which is an alarm means for alarming the abnormality to the outside when the detected person in the bathroom 2 has an abnormality. 4b and a release switch 4c for releasing the alarm means of the loudspeaker 4b. In the loudspeaker 4b, the sound pressure level value of the alarm sound is set to at least two levels, a small level and a large level.

Next, in the operation panel 4, based on the sound and the signal input relating to the movement of the body input from the microphone 3a and the infrared distance measuring sensor 3b,
A control device 5 for controlling the alarm 3c and the loudspeaker 4b is provided. The control device 5 is a well-known device that has a microcomputer or the like therein, and has a CPU, ROM, RAM, etc., which are not shown in the microcomputer, and is operated by the operation input of the operation switch 4a of the operation panel 4. Is.

Next, the control device shown in FIG. 2 will be used as a processing procedure when an abnormality occurs such as when the detected person stops moving due to cardiac arrest in the bathroom 2 which is the main part of the present invention. A description will be given based on the control flowchart of No. 5. First, before the detected person enters the bathroom 2, the operation switch 4
When a is turned on and power is supplied to the control device 5, the main routine shown in FIG. 2 is started. And step 10
At 0, the contents stored in the data processing memory (RAM) are initialized. Step 1 which is the next occupancy determination means
In 10, the determination is made by detecting whether or not the detected person has entered the bathroom 2 by the infrared distance measuring sensor 3b, and at this time, the person has not entered the bathroom 2 yet. Therefore, it is determined to be NO and step 12
Move to 0.

In step 120, the background noise storage means (storage means) collects the acoustic waveform of the background noise in the bathroom 2 when no person is present in the bathroom 2 from the microphone 3a and stores it in the RAM. . The detected person is in the bathroom 2
When the user enters the room, YES is determined by the occupancy determination unit in step 110, and the process proceeds to the detection start in step 130.

In this step 130, the sound waveform generated by the detected person moving the body in the bathroom 2 and the movement of the detected person's body are detected by the microphone 3a and the infrared distance measuring sensor 3b and the detected waveform. By inputting the signal of the distance to the person, the monitoring of the detected person is started.

Here, the monitored acoustic waveform and the distance to the subject will be described with reference to FIG. First, the control device 5 has a signal processing unit (not shown), and inputs a signal of the acoustic waveform and the distance to the detected person from the microphone 3a or the infrared distance measuring sensor 3b.
In the case of the acoustic waveform signal, the following 3 depending on the motion of the detected person.
Acoustic waveforms of sound pressure characteristics, which are roughly classified into types, are input and monitored.

That is, the first one is the background noise characteristic X representing the background noise in the bathroom 2 before the person to be detected enters the bathroom 2 described above. The background noise characteristic X becomes a reference value for determination in step 140 described later, and the sound pressure characteristic is a low value as shown in FIG. 3 because the sound source is outside the bathroom 2.

Next, there is a cleaning sound characteristic Y which represents a cleaning sound generated when the person to be detected is cleaning his / her body in the washing room 2b. And the third is a bathing sound characteristic Z that represents a bathing sound generated while the subject is bathing in the bathtub 2c. In general, the cleaning sound characteristic Y has a larger sound pressure characteristic than the bathing sound characteristic Z. The background noise characteristic X is stored in the RAM in step 120 described above.

Further, the distance signal is used to obtain the moving distance characteristic XX in which the detected person moves his / her body in the bathroom 2.
For example, the distance from the entrance to the bathroom 2 to the exit from the bathroom 2 via the washing room 2b and the bathtub 2c is monitored by displaying the moving distance per unit time by the distance between the detected person and the infrared distance measuring sensor 3b.

In these monitor signals, the signal showing the acoustic waveform is the first information, and the signal showing the moving distance is the second information.
Information and. Then, for example, when an unexpected accident such as cardiac arrest occurs in the detected person in the bathroom 2, the movement of the detected person is stopped, and the movement distance and the acoustic waveform change. That is, the stopped state is continued at the moving distance, and the acoustic waveform is maintained in the silent region state in which the sound pressure characteristic which is almost similar to the background noise characteristic X is low level.

Therefore, in the next step 140 which is the first judging means, the background noise characteristic X stored in step 120 is compared with the sound pressure characteristic in the monitored silent region state, and the sound pressure characteristic is less than a predetermined amount. And the first predetermined time (T
It is a first abnormality determination means for determining whether or not 1) has elapsed to determine the abnormal operation state of the detected person. If YES here, the process proceeds to step 150.

This step 150 is a notification means for inquiring of the person to be detected whether or not there is any abnormality. For example, an inquiry is made by a voice signal such as "Is it okay?" Or "What happened?" Is. Note that this audio signal is stored in advance in a ROM (not shown),
The information is output to the alarm device 3c to interrogate the subject.

Then, at step 160, the response determining means for determining whether or not there is a response to the inquiry, determines whether or not there is an abnormality in the subject. Specifically, the subject makes a voice to change the sound pressure characteristic of the silent region, and it is determined whether or not the sound pressure characteristic of the silent region or more is detected and there is a response from the detector. If there is a response, the process proceeds to step 140 again. Note that this step 14
In the first abnormality determination means of 0, the first abnormality determination means is released when the sound pressure characteristic of "response" continues for a predetermined time. This is so that when the sound pressure characteristic of “with response” is an extremely short response, the first abnormality determination means is not released and the notification means of step 150 is checked again.

Then, when there is no response in the response determining means in step 160, the next second determining means, step 170, is executed. Here, the moving distance characteristic X
It is a second abnormality determination means that determines whether or not the stopped state of the detected person is continued from the monitor of X and determines the abnormal operation state of the detected person. Then, if the signal is a stop state, the process proceeds to step 180.

This step 180 is an alarm means, and more specifically, a means for alarming a person in the living room in which the operation panel 4 is arranged that the detector has an abnormality in the bathroom 2. Then, the control device 5 controls the loudspeaker 4b to output an output signal of a small level, and the loudspeaker 4b emits an alarm sound.

Then, in addition to issuing an alarm, in the next step 190, in order to confirm the situation in the bathroom 2, whether or not a change has occurred in either the acoustic waveform of the monitored signal or the signal of the moving distance. judge. Then, if there is no change in any of these characteristics, the process proceeds to step 200 which is the next alarm means. If any of these characteristics change, there is a change.
Then, the process proceeds to step 140.

The alarm means of step 200 is to issue an alarm sound from the loudspeaker 4b at a high level, which is a higher level than that of the alarm means of step 180 described above. The release switch 4 is used to release this alarm means.
The alarm sound is released by activating c. Therefore, the alarm sound continues until the release switch 4c is released in step 210.

The above-mentioned processing procedure has been mainly described for the abnormality detection, the notification means and the warning means when the silent area is reached in the motion of the detected person and the motion continues to be stopped.
When an abnormality occurs in the person to be detected while continuing to take out the shower, the faucet, etc., since the acoustic waveform exceeds the low level range, the first abnormality determining means in step 140 determines NO. At this time, the determination is made by the second abnormality determining means in step 170 described above. That is,
If the movement distance characteristic XX of the person to be detected is a signal indicating that there is no abnormality if the stopped state does not continue, and the signal indicates the stopped state, the process proceeds to step 220. This prevents the following third abnormality determining means from being executed when the body of the subject is moving.

Then, step 22 which is the third judging means.
At 0, the acoustic waveform corresponding to the middle level having a larger sound pressure characteristic than the stored background noise characteristic X is monitored, and the sound pressure characteristic in the middle sound region is monitored and the second predetermined time (T2) is reached. It is a third abnormality determination means that determines whether or not the time has elapsed and determines the abnormal operation state of the subject. Then, if the second predetermined time (T2) has elapsed and the result is YES, the process proceeds to step 150.

This step 150 is a notification means for asking the person to be detected whether there is any abnormality, similar to the above-mentioned step 150, and here, for example, "No water is coming out!" The question is asked by a voice signal such as "?". The voice signal is stored in advance in a ROM (not shown), and is output to the alarm device 3c to inquire of the person to be detected.

Then, at step 160, the response determining means for determining whether or not there is a response to the inquiry, determines whether or not there is an abnormality in the subject. Specifically, whether the detected person stops the shower, faucet, etc. and changes the sound pressure characteristic such as the sound pressure characteristic in the middle sound range is lowered,
Alternatively, it is determined that there is a response by changing the movement of the body and changing the movement distance characteristic XX. If there is a response, the process proceeds to step 220 again via step 140. In the third abnormality determining means of this step 220, the third abnormality determining means is released when the sound pressure characteristic of “with response” continues for a predetermined time or when the moving distance characteristic XX changes from the stopped state. It is supposed to be done. This is because when the characteristic of “with response” is an extremely short response, the third abnormality determination means is not released, and step 150
I am trying to check the notification means again.

Then, the response determining means at step 160 determines that there is an abnormality when there is no response, and executes the alarm means at step 200.

In step 210, the alarm sound is released by releasing the release switch 4c. However, the alarm sound is changed and the characteristics of the moving distance characteristic XX are changed so that the alarm sound is released after continuing for a predetermined time. Is also good. Further, the microphone 3a preferably has a detection accuracy capable of sensitively capturing a sound of water when using a shower, a faucet, or the like in a frequency band, a waveform, and the like.

According to the human body abnormality detecting device 1 of the above embodiment, the microphone 3a and the infrared distance measuring sensor 3b are installed in the bathroom 2, and the acoustic waveform which is the first information indicating the sound generated by the person to be detected. And first and second abnormality determining means steps 140 and 170 for determining the abnormal operation state of the detected person based on the second information indicating the movement of the detected person calculated from the distance to the detected person. As a result, compared to a conventional monitoring device using a camera, for example, in sound, a reverberant sound is louder in a closed space or a narrow space, and a sound that enters from outside the detection space propagates in the air. It is easy to detect the presence or absence of sound because it is easily attenuated by passing through the wall. Especially when it is applied to a dark room or a bathroom where steam is generated, the detection accuracy is superior to the image using a camera.

Further, in a camera, a narrower space restricts the image range to be detected. Therefore, the mounting position must be restricted or the camera must have an additional function such as a tracking means. There is no restriction on the mounting position and there is no need to follow the directivity.

Further, it is possible to provide a device capable of detecting accidental accidents such as cardiac arrest while taking a bath, without fear of causing sexual dissatisfaction as compared with the conventional video system.

Further, by monitoring the movement of the person to be detected in the bathroom 2 with the acoustic waveform and the moving distance, when either the silent area or the middle sound area is detected from those monitor signals, the alarm means is provided. By controlling the loudspeaker 4b in steps 180 and 200, an echo effect due to a larger echo is obtained in a narrower private room, particularly in an acoustic waveform, so that detection can be performed with higher accuracy than in the video system.

In order to determine the abnormal operation state, the first abnormality determining means step 140 for determining a silent area, the third abnormality determining means step 220 for determining a middle tone area, and the second abnormality determination for determining the moving distance. Means step 170
By having and, the sound pressure characteristic of the silent region when the shower or the faucet is not used and the sound pressure characteristic of the medium sound region when the shower or the faucet are used can be detected separately. ,
Then, after confirming with the person to be detected by the notification means step 150, it is first determined with the response determining means step 160 whether or not there is a response, thereby first performing mutual confirmation with the person to be detected existing in the bathroom 2. Therefore, the stepwise human body abnormality detection device 1 having a dialogue with the person to be detected becomes possible.

Further, by comparing the background noise storage means step 120 storing the monitored first information as a reference value, it is easy to determine the abnormal operation state.

(Second Embodiment) In the above embodiments, the sound generated in the bathroom 2 is detected by the microphone 3a, the sound waveform is monitored as it is, and compared with the background noise characteristic X to detect the abnormal operation of the person to be detected. I explained the judgment of the state,
Not limited thereto, the sound generated in the bath 2 is detected by the microphone 3a, the specific frequency component by a band pass filter, for example 2000~4000H _Z, 300
0H _Z, may detect the presence or absence of sound absolute level of that component by extracting such 4000H _Z.

Further, the microphone 3a provided in the remote control panel 3, the alarm 3c, and the loudspeaker 4b provided in the operation panel 4 are used for inputting or outputting voice, and the operation panel 4 also has a bathroom. 2 Provide a microphone for inputting voice from outside and a call switch,
It is preferable to configure a communication means between the inside of the bathroom 2 which is the detection space and the outside of the bathroom 2.

Specifically, as shown in FIG. 4, in the present embodiment, the second microphone 4d and the call switch 4e are provided in the operation panel 4, and the control device 5 is provided with the first band which is the first filter means. Pass filter 5a
And a second bandpass filter 5b, which is a second filter means.

The first bandpass filter 5a is a filter for extracting only the above-mentioned specific frequency component from the sound input through the microphone 3a. It is advisable to extract at least two or more components of the specific frequency component to be extracted, or to extract at the same time, and to detect the presence or absence of a predetermined sound by the absolute level of the extracted frequency component.

The second band pass filter 5b is
When a call is made between the detected person and the outside of the bathroom 2, the detected person's voice input to the microphone 3a is, for example, 200
A filter for removing or attenuating a specific frequency component of 0～4000H _Z, thereby, cracking sound and rasp due reverberation or resonance sound in the bathroom 2 is to remove or attenuated by clear speech be able to.

Here, with regard to the extraction of the specific frequency component by the first filter means 5a in the present embodiment, the representative sound among the sound entering the bathroom 2 and the sound generated in the bathroom 2 is analyzed. Or, it demonstrates based on FIG. Fig. 5 is the sound that the hot water moves when you move your body in the bathtub, and Fig. 6 is the sound when the vehicle is running outside the bathroom. Fig. 7 is the door that is closed close to the bathroom 2. 8 is the sound of the shower pouring toward the washroom by hooking the shower head on the rack, and FIG. 9 is the sound when the body is taking a shower with the shower head in hand. In each figure, (a) is a characteristic diagram of the amplitude fluctuation with respect to the elapsed time (second) detected by the microphone 3a, and (b) is the locus of the peak value for 0.5 seconds when the sound is subjected to high-speed frequency conversion. A characteristic diagram, (c) shows the sound detected by the microphone 3a in the bandpass filter 5
characteristic diagram of amplitude variation with respect to the elapsed time of extracting specific frequency components of 2600～4000H _Z by a (second)
(D) is an enlarged characteristic diagram of the portion indicated by the ellipse in (c).

First, referring to FIG. 5 and FIG. 6, comparing the characteristic diagrams of the amplitude variation of FIG. 5A, there is no great difference in sound pressure level, and it is difficult to discriminate in this state. Therefore, it can be seen that attenuated sound pressure level in it is 500H _Z vicinity or sound Figure 6 when compared with high-speed frequency conversion as shown in (b). Next, when a specific frequency component is extracted by the bandpass filter 5a according to the present invention and compared, as shown in (c) and (d) of FIG.
The sound pressure level is obviously higher than that of FIG.

Therefore, as shown in the figure, the first threshold value (L
By setting 1) and the third predetermined time (T3), the sound intruding from the outside in FIG. 6 cannot satisfy the first threshold (L1). This makes it possible to detect the sound generated by the movement of the person in the bathroom 2 without being disturbed by the sound that enters from outside the bathroom 2.

Next, the sound when the door is violently closed near the bathroom 2 shown in FIG. 7 will be described.
Since the amplitude immediately after the sound is generated is larger than that of other sounds, the value of the sound pressure level on the vertical axis is reduced to about 1/5. Then, when the specific frequency component of the sound is extracted by the bandpass filter 5a according to the present invention, as shown in FIG. 7 (d), there are instantaneously several parts that exceed the first threshold (L1). However, at the same time, the third predetermined time (T3) is not satisfied. Even if the third predetermined time (T3) is satisfied, as will be described later, if this detection sound is not detected again within a predetermined time range (T4 to T5) thereafter,
By determining that the detected sound detected earlier is invalid, the presence or absence of sound can be detected more reliably.

Next, FIGS. 8 and 9 both analyze the sound when the shower is opened, and FIG. 8 which is the shower sound when the shower head is hooked on the rack and pouring toward the washroom is shown in FIG. It is monotonous and its sound pressure level is relatively low.
Almost no part exceeds the threshold (L1) and the third predetermined time (T3). On the other hand, in the sound when the body is taking a shower while holding the shower head in FIG. 9, the sound pressure level constantly changes, and the first threshold value (L1) and the third predetermined time (T3) are sufficiently satisfied. It is a sound that can be reliably detected.

In this example, the sound that can be detected in the bathroom is set by the first threshold value (L1) and the third predetermined time (T3) based on the typical sound in the bathtub. Not limited to this, whether a hot water supply device such as a hot water supply device (not shown) that supplies hot water to the bathroom causes the control device 5 to input a hot water supply operation signal that is hot water supply operation information indicating that the hot water supply device is operating or stopped to supply hot water to the bathroom. Whether or not the shower or the faucet is left open may be determined by determining whether or not the shower or the faucet is left open.

At this time, the second threshold value (L1) and the sixth predetermined time period (T6) are changed to the second threshold value (L1) and the sixth predetermined time period (T6) instead of the first threshold value (L1) and the third predetermined time period (T3).

Although the sound detected by the microphone 3a has been described as a waveform as it is here, in general, when the presence or absence of sound is electrically processed, the AC component is once converted into a direct current by an AC / DC inverter. It is converted into a component, and it is determined whether or not the threshold value is exceeded at the rising of the DC component, and whether or not the predetermined time can be maintained by the falling time. The present invention does not refer to such determination means.

However, as a concrete judgment means, AC / D
In the case of the determination by the C conversion, it has been found that it is possible to more reliably determine whether the shower or the faucet remains open by setting the time constant at the time of conversion as follows. e), FIG. 8E, and FIG. 8F will be described.

In FIGS. 5 (e) and 8 (e), the + side of the waveform shown in (d) is half-wave rectified, and AC / DC conversion is performed with a time constant of 2 msec, which is the first time constant. A range in which the direct current component, the fixed threshold value and the combined threshold value obtained by adding the variable threshold value AC / DC converted with the time constant 12 msec which is the second time constant longer than the first time constant and the direct current component exceeds the combined threshold value Is.

Further, FIG. 8 (f) shows a range in which the direct current component exceeds the composite threshold value by obtaining the composite threshold value in the same manner as in (e) above in the part indicated by the double line ellipse in (c). Is shown.

First, referring to FIG. 5 (e), when a sound that is emitted as a chap is input, the DC component and the variation threshold value start increasing at their respective time constants. At this time, since the variation threshold has a large time constant, it rises with a delay from the rise of the DC component, so that the DC component immediately exceeds the synthesis threshold.

A direct current component having a small time constant begins to fall gently as the input voice drops, but the variation threshold has a large time constant, and thus remains stable and remains almost unchanged. Therefore, the range in which the DC component exceeds the synthesis threshold is the seventh predetermined time (T7) or more, and it is determined that the bather is operating.

Also in FIG. 8 (e) in which the shower head is hooked on the rack and the sound of the shower pouring toward the washing room is input, when the shower is started to start, the DC component exceeds the synthesis threshold value. The range is longer than the seventh predetermined time (T7), and it is determined that the bather is operating.

However, in FIG. 8 (f) in which the shower is left open, the waveform B before the AC / DC conversion is performed.
The part is about the same as the waveform A part shown in FIG. 8 (e), but the fluctuation threshold value is stably higher than that of the continuous shower sound, and a large voice is temporarily input. However, the range in which the DC component exceeds the synthesis threshold is less than the seventh predetermined time (T7), and it can be determined that the bather is not operating without being detected.

As a result, when the shower or faucet is left open, only the start of the discharge is detected and the state of the rest after that is not detected. Therefore, the bather does not operate during the standing. , Can be determined more reliably.

Next, the control processing for determining the abnormal operation state of the detected person according to the second embodiment will be described based on the control flowchart of the control device 5 shown in FIG.

First, the monitor starts (step 130).
Then, in step 310, a second timer (for example, about 20 to 30 seconds) that does not detect a single sound that closes the door or the like, and a third timer (T5 that counts the generation interval of the detected sound). , And a fifth timer (T1, for example, about 1 minute) for counting the silent state until the subject is asked whether there is any abnormality.

In the next step 320, the hot water supply operation information determining means determines whether or not a hot water supply device (not shown) is operating to stop hot water in the bathroom, and there is a hot water supply signal. If so, the hot water supply device is operating, and if there is no hot water supply signal, the hot water supply device is stopped.

If there is no hot water supply signal, the routine proceeds to step 330, where the first threshold value (L) for judging the presence or absence of sound after extraction of the specific frequency component by the bandpass filter 5a.
1) and a third predetermined time (T3) for setting the time of the first timer for sound duration determination. If there is a hot water supply signal, in step 340, the first threshold value (L
The second threshold value (L1) and the sixth predetermined time period (T6) that are larger than 1) and the third predetermined time period (T3) are set.

Next, step 140 of the first abnormality determining means, which is the first determining means for determining the abnormal operation state of the person to be detected.
a, and the set first or second threshold value (L
1, L2) or more is determined. And if it is more than the 1st or 2nd threshold value (L1, L2), bathroom 2
The process proceeds to the acoustic determination routine 400 for determining whether or not the detected sound is the action sound of the detected person. On the contrary, if it is less than or equal to the first or second threshold value (L1, L2), the process routine 500 at the time of abnormality is performed.

When the detected sound reaches the first or second threshold value (L1, L2) or more, the first timer for duration determination for confirming the duration of this sound starts counting in step 410. To be done. Next, step 42.
At 0, it is determined whether or not the first timer has timed up the set third or sixth predetermined time (T3 or T6). If the time is not up, step 3
The process returns to 10 and the sound detection is repeated again. On the other hand, when the time is up, the routine proceeds to step 430, where it is determined by the second timer whether or not the sporadic sound is invalidated.

Here, when the second timer has not timed up, that is, when the silent time is relatively short, the action in the bathroom is active, and the detected person in the bathroom should be active. Even if a part of the sound that does not accompany the motion is detected, there is no problem from the human body abnormality detection. Therefore, the process proceeds to step 450, and the second, fifth, sixth and seventh timers are reset. That is, it is set here that the detected person is in action.

If the second timer times out in step 430, it means that the detected person has stopped operating for a while. Therefore, if the stop of the operation is due to some accident, for example, if a sound from the outside is detected and it is determined that the person to be detected is in operation, the alarm is delayed. Therefore, in order to eliminate the sporadic sound that is accidentally detected for some reason, it is reconfirmed in the next step 440.

That is, at step 440, the third timer causes the first or second threshold value (L1, L2) or more, and the third or sixth predetermined time (T3 or T6).
The time since the last detection sound was detected, that is, the generation interval of the detection sound is determined.

If the third timer has timed out, step 450 is not executed as a single sound (the second, fifth, sixth and seventh timers are not reset),
In other words, it is determined that the detected person is not operating, and only the first and third timers in step 460 are reset.

Then steps 470, 480 and 490.
Move to. In these steps, when the detection sound of the first or second threshold value (L1, L2) or more and the detection sound of the third or sixth predetermined time (T3 or T6) or more is detected once,
The next acoustic detection is stopped for the fourth predetermined time (T4, for example, about 1 second). This results in a plurality of acoustic groups having several peak values even if the single sound when the door is hard closed as shown in FIG. Run to catch as sound.
Then, the process returns to step 310 and the sound monitoring is repeated.

Next, the processing routine 500 at the time of abnormality will be described. First, in step 510, the first timer for duration determination is reset. Next, the process proceeds to step 140b of the first abnormality determining means, which is the first determining means, and the duration of the silent state is confirmed by the fifth timer. If the fifth timer has not timed out, it is determined that the abnormal state has not been reached, the process returns to step 310, and the sound monitoring is repeated.

On the other hand, when the fifth timer has timed out, the process moves to the next notification means, step 150a, and a voice signal such as "Is it okay?" Or "What happened?" Inquires about the presence or absence of abnormality in the detected person, and waits for the response time to the inquiry No. 6
Start timer counting. Since this inquiry is performed intermittently, sound detection may be interrupted during the inquiry, and the response from the detected person may be determined between the questions.

In the present embodiment, the voice signal used in this inquiry is the first bandpass filter 5a.
The specific frequency component extracted by is previously removed or attenuated. Accordingly, the inquiring voice informed in the bathroom is not erroneously determined as the operation sound of the detected person, and the accurate first information can be obtained even during the informing.

Next, the process proceeds to step 160a which is a response means, and it is determined whether or not there is a response from the detected person. If there is a response and the sixth timer has not timed out yet, the process returns to step 310 and the sound monitoring is repeated. Further, if there is a response here, it is determined to be normal by the acoustic determination routine 400 described above.

Then, when the sixth timer times out, it is determined that there is no response from the detected person in the sixth predetermined time (for example, about ten and several seconds), and the process proceeds to step 180 which is an alarm means. In step 180, the detected person is continuously inquired, and the person outside the bathroom 2 is warned that the detected person has an abnormality. This alarm is an output signal of a small level as an initial stage and the loudspeaker 4b is used.
Is a warning sound. Further, at the same time, the counting of the judgment waiting seventh timer (for example, about ten and several seconds) for judging whether or not the treatment by this alarm is performed is started.

Then, when the seventh timer in step 190a has timed out, it is determined that no action has been taken for the abnormality of the detected person, and the next alarm means, step 2
00. In step 200, while continuously inquiring about the detected person, the fact that an abnormality has occurred in the detected person in the room outside the bathroom 2 continues to be output with an output signal of a level higher than that in step 180, and then step 210 Move to. In step 210, it is determined whether or not a release operation has been performed with the release switch 4c. If there is no release operation, the process returns to step 200 to continue issuing an alarm, and if the release operation is performed, the process returns to step 100.

According to the second embodiment described above, the presence or absence of sound in the bathroom 2 can be detected more accurately and easily by extracting the detected sound into the specific frequency component by the first bandpass filter 5a. . In addition, by switching at least two or more specific frequency components from the sound generated in the bathroom 2 and extracting them at the same time, the resonance or the resonance frequency changes due to the change in the size of the bathroom 2. The sound generated in the bathroom 2 can be reliably detected.

Further, by extracting only the sound that is relatively loud and has a long duration from the sounds generated in the bathroom 2, the sound that enters from outside the bathroom 2 is extracted by the first band pass filter 5a. Even if the specific frequency component is not completely attenuated and remains, the sound in the bathroom 2 can be accurately and easily detected.

In the dressing room or washroom adjacent to the bathroom 2, a single sound that closes the door or the door of the closet with vigor is judged to be in a substantially silent state unless reproduced thereafter. Even if a single sound from the outside of the bathroom 2 is detected, it is determined that the sound is substantially silent, so that erroneous detection can be prevented when determining the abnormal operation state of the subject.

Further, the hot water supply operation information (step 320) indicating the operation of the hot water supply device is provided in the control device 5, and specifically, depending on the presence / absence of the hot water supply signal, specifically, the person to be detected supplies hot water in the bathtub. Or when you use the hot water supply by opening the faucet or shower such as taking a shower at the washroom,
Since the sound inside is different, by changing to the second threshold value (L2) and / or the second predetermined time (T6), even when the faucet or the shower is kept open, Since it is superimposed on repetitive sounds or repetitive vibrations that do not involve the subject's continuous motion, only the motion noise or vibration of the subject is detected to reliably determine the abnormal motion state of the subject. be able to.

Further, by removing or attenuating the specific frequency component extracted by the first band pass filter 5a in advance from the voice signal by the notifying means (step 150a), only the operation sound of the subject is notified even during the notification. Can be accurately detected.

(Other Embodiments) In the above embodiments, the judging means for judging the abnormal operation state by detecting the sound with the microphone 3a and monitoring the acoustic waveform as it is or extracting it into the specific frequency component. However, the present invention is not limited to this, and a vibration pickup or the like that detects vibration in the bathroom 2 may be used to detect vibration and monitor the vibration waveform to determine the abnormal operation state. In a bathroom such as a bath unit where a washing room and a bathtub are integrally formed, when a vibration waveform is used to determine an abnormal operation state, a vibration pickup can be installed on the wall surface where vibration noise is easily transmitted to detect vibration. Is improved.

Further, in the first embodiment, the description is provided including the determination means of step 170 of the second abnormality determination means which is the second determination means after the determination that there is no response which is the response determination means. However, the present invention is not limited to this. First, the response determination means (step 160) after the acoustic waveform has passed the silent region for the first predetermined time (T1).
If there is no response in step 170 of the second abnormality determination means
May be omitted and the next alarm means may be executed. As a result, it is possible to promptly execute the alarm means.

Further, in the above embodiment, the steps 150 and 150a, which are the notification means, are queried with respect to the detected person in the bathroom 2 by the voice signal, but it may be notified as a warning sound. Moreover, the warning sound may be generated outside the bathroom 2. As a result, it is possible to take precautions as a prediction warning of an abnormality to a person existing outside the bathroom 2.

Further, in the above embodiment, the infrared distance measuring sensor 3b which is the distance measuring means for detecting the distance to the detected person is used as the room entrance detecting means for detecting the entering / leaving of the detected person into / from the bathroom 2. However, the detection of the entering / leaving room of the person to be detected is not limited to this, but the detection start manual switch, the opening / closing signal of the bathroom door 2a, the passage signal of the detected person, the lighting switch of the bathroom 2 and the like are detected. You can also do it. Further, in the bathroom 2, it is also possible to detect the entering and leaving of the person to be detected by detecting a start signal of running water such as a shower or a hot water tap in the bathroom.

Also, the voice of the person to be detected, which is input to the microphone 3a, is output from the loudspeaker 4b after the specific frequency component is removed or attenuated by the second bandpass filter 5b, and is reflected in the bathroom 2. It is possible to remove or attenuate the sound cracking or the offensive sound due to the sound or the resonance sound to obtain a clear voice.

Further, although the microphone 3a, the infrared distance measuring sensor 3b, and the alarm 3c are arranged in the remote control panel 3, they may be incorporated in a remote control panel such as a hot water supply device for supplying hot water to the washing room 2b or the bathtub 2c. . Thereby, the additional function of the water heater is improved.

Further, although the human body abnormality detecting device 1 applied to the bathroom 2 has been described in the above embodiment, the present invention is not limited to this, and the detection space for detecting the human body abnormality according to the present invention is a washroom, It may be applied to a toilet or a living room for a detected person who needs care such as an elderly room.

[Brief description of drawings]

FIG. 1 is a mounting explanatory diagram showing an overall configuration of a human body abnormality detection device 1 according to a first embodiment of the present invention.

FIG. 2 is a flowchart showing a control flow of a control device 5 in the first embodiment of the present invention.

FIG. 3 is a characteristic diagram showing acoustic waveform characteristics and movement distance characteristics of the signal processing unit according to the first embodiment of the present invention.

FIG. 4 is an installation explanatory diagram showing the overall configuration of a human body abnormality detection device according to a second embodiment of the present invention.

5 (a) is a characteristic diagram of amplitude fluctuation with respect to elapsed time (seconds) detected by the microphone 3a, and FIG. 5 (b) is a peak value when high-speed frequency conversion is performed on the sound. characteristic diagram of the trajectory, (c) is a characteristic diagram of amplitude variation with respect to the elapsed time of extracting specific frequency components of 2600～4000H _Z by the first band-pass filter 5a sounds detected by the microphone 3a (in seconds), (d) Is an enlarged characteristic diagram of the portion indicated by the ellipse in (c),
(E) is a characteristic diagram showing a range in which a direct current component when the + side of (d) is half-wave rectified and subjected to AC / DC conversion exceeds a synthesis threshold.

FIG. 6A is a characteristic diagram of the amplitude variation with respect to the elapsed time (seconds) detected by the microphone 3a, and FIG. 6B is a peak when the high-speed frequency conversion of the traveling sound of the vehicle emitted outside the bathroom 2 is performed. characteristic diagram of the trajectory values (c) are characteristic diagram of amplitude variation with respect to the elapsed time of extracting specific frequency components of 2600～4000H _Z by the first band-pass filter 5a sounds detected by the microphone 3a (in seconds), ( FIG. 3D is a characteristic diagram in which the portion indicated by the ellipse in FIG.

7 (a) is a characteristic diagram of the amplitude fluctuation with respect to the elapsed time (seconds) detected by the microphone 3a, and FIG. 7 (b) is a peak when the sound is fast-frequency converted. A characteristic diagram of the locus of values, (c) shows the sound detected by the microphone 3a in the first bandpass filter 5
characteristic diagram of amplitude variation with respect to the elapsed time of extracting specific frequency components of 2600～4000H _Z by a (second)
(D) is a characteristic diagram in which the portion indicated by the ellipse in (c) is enlarged.

8A is a characteristic diagram of the amplitude variation with respect to the elapsed time (seconds) detected by the microphone 3a, and FIG. 8B is a peak value when the sound is subjected to high-speed frequency conversion. characteristic diagram of the trajectory, (c) is a characteristic diagram of amplitude variation with respect to the elapsed time of extracting specific frequency components of 2600～4000H _Z by the first band-pass filter 5a sounds detected by the microphone 3a (in seconds), (d) Is an enlarged characteristic diagram of the portion indicated by the ellipse in (c),
(E) is a characteristic diagram showing a range in which the direct current component when the + side of (d) is half-wave rectified and subjected to AC / DC conversion exceeds the synthesis threshold, and (f) is the double line in (c). FIG. 9 is a characteristic diagram showing a range in which a direct current component when AC / DC conversion is performed on a portion indicated by an ellipse exceeds a synthesis threshold.

[Fig. 9] A sound emitted from a shower in the bathroom 2 (a)
Is a characteristic diagram of amplitude fluctuation with respect to elapsed time (seconds) detected by the microphone 3a, (b) is a characteristic diagram of a locus of peak values when the sound is subjected to high-speed frequency conversion, and (c) is detected by the microphone 3a. characteristic diagram of amplitude variation with respect to sounds an elapsed time of extracting specific frequency components of 2600～4000H _Z by the first band-pass filter 5a (s),
(D) is a characteristic diagram in which the portion indicated by the ellipse in (c) is enlarged.

FIG. 10 is a flowchart showing a control flow of the control device 5 in the second embodiment of the present invention.

[Explanation of symbols]

3a ... Microphone (detection means) 3b ... Infrared distance measuring sensor (entrance detection means, distance measurement means) 3c ... Alarm device 4b ... Loudspeaker 4d ... Second microphone 5a ... First band pass filter (first filter means) 5b ... 2nd band pass filter (2nd filter means) L1, L2 ... 1st threshold value, 2nd threshold value 120 ... Background noise storage means (storage means) 140, 140a ... 1st abnormality determination means (1st determination means) 150, 150a ... informing means 160 ... response judging means 170 ... second abnormality judging means (second judging means) 180, 200 ... warning means

Continued front page    (72) Inventor Takahiro Mita             1-1, Showa-cho, Kariya city, Aichi stock market             Inside the company DENSO F-term (reference) 5C086 AA22 BA04 CA09 CA12 CA23                       CB16 CB26 DA08 DA40 EA15                       EA17 EA31 FA06 FA11

Claims (16)

[Claims] 1. A detection means (3a) for detecting a sound or a vibration in a detection space, an entrance detection means (3b) for detecting the entrance of a person to be detected into the detection space, and the entrance detection means (3b). ) Detects a person to be detected and starts a monitoring operation, and first determines an abnormal operation state of the person to be detected based on first information indicating sound or vibration detected by the detecting means (3a). Judgment means (140, 140
a) A human body abnormality detecting device comprising: 2. The human body abnormality detecting device according to claim 1, wherein the entrance detecting means (3b) is a distance measuring means (3b) for detecting a distance to a person to be detected. 3. The first determining means (140, 140a)
2. The abnormal state of the detected person is determined by detecting that the first information indicating sound or vibration is in a substantially silent state or a non-vibrated state for a predetermined time. The human body abnormality detection device according to claim 2. 4. The first determining means (140, 140a)
Is a second determination for determining the second information indicating the movement of the detected person from the distance information detected by the distance measuring means (3b), and for determining the abnormal operation state of the detected person based on the second information. Means (170), the second determination means (170) detects that the second determination means (170) does not substantially change based on the second information, and the first information indicating sound or vibration is in a substantially silent state for a predetermined time or The human body abnormality detection device according to claim 3, wherein the abnormal state of the detected person is determined by detecting a non-vibration state. 5. The first determining means (140, 140a)
Stores the first information indicating the sound or the vibration detected by the detection means (3a) when the detected person is not detected by the room entrance detection means (3b) as background noise information or background vibration information. The background noise information or the background vibration information is used as a reference value when the storage means (120) is provided, and the reference value at the time of detecting the substantially silent state or the vibration free state is used. Item 5. The human body abnormality detection device according to any one of items 4. 6. The first determining means (140, 140a)
Is the first filter means (5a) based on the first information indicating the sound or vibration detected by the detection means (3a).
6. The specific frequency component is extracted by the method, and it is detected that there is a substantially silent state or a non-vibrating state, and the abnormal operation state of the detected person is determined.
The human body abnormality detection device according to any one of 1. 7. The first filter means (5a) is configured to switch and extract at least two or more components as specific frequency components to be extracted, or simultaneously extract. 6. The human body abnormality detection device according to 6. 8. The first determining means (140, 140a)
Is a predetermined first threshold value (L1) or more among the specific frequency components extracted by the first filter means (5a),
Further, a sound or vibration continued for a third predetermined time (T3) or more is used as a detection sound or a detected vibration, and it is detected that there is no substantial silent state or no vibration state, and the abnormal operation state of the detected person is determined. Claim 6 or claim 7
The human body abnormality detection device described in. 9. The first determination means (140, 140a)
Even if a sound or a vibration continued for a predetermined first threshold value (L1) or more and a third predetermined time (T3) or more is detected as the detected sound or the detected vibration, a predetermined range time (T4) after that is detected. ~ T5), when the detected sound or the detected vibration is not detected again, it is detected that the state is substantially silent or no vibration, and the abnormal operation state of the subject is determined. The human body abnormality detection device according to claim 8. 10. The detection space is a bathroom, and a hot water supply device for supplying hot water into the bathroom and an operation of the hot water supply device,
When the hot water supply operation information indicating the stop is provided and the hot water supply information indicating the operation of the hot water supply device is input, the first determination means (140, 140a) causes the predetermined first threshold value (L1) or / And the third predetermined time (T3) is changed to the second threshold value (L2) and / or the sixth predetermined time (T6), and it is detected that there is substantially no sound or no vibration, and the abnormality of the detected person is detected. The human body abnormality detection device according to claim 8 or 9, wherein the operating state is determined. 11. The first determining means (140, 140)
a) shows the specific frequency component extracted by the first filter means (5a) as AC / DC with a predetermined first time constant.
While converting into a DC component, AC / DC conversion is performed with a second time constant longer than the first time constant as a variation threshold,
A fixed threshold value is added to this, and as a combined threshold value, when the DC component exceeds the combined threshold value for a seventh predetermined time (T7) or more, it is determined that it is not in a substantially silent state or a non-vibrating state. The human body abnormality detecting device according to any one of claims 1 to 10. 12. The first determining means (140, 140)
a) includes notifying means (150, 150a) for notifying the detected person when the abnormal operation state of the detected person is determined, and the first determination means (140, 140a) is the abnormal operation state of the detected person. When it is determined that the notification means (150, 1
The human body according to any one of claims 1 to 11, further comprising a response determination means (160) for determining whether or not there is any abnormality in the detected person after the notification is given to the detected person by 50a). Anomaly detection device. 13. The voice output from the notification means (150, 150a) is the first filter means (5a).
13. The human body abnormality detecting device according to claim 12, wherein the specific frequency component extracted by is a voice that has been removed in advance. 14. The first determining means (140, 140)
The alarm means (180, 200) for outputting an alarm to the outside of the bathroom when the abnormal operation state of the detected person in the detection space is determined according to a).
The human body abnormality detection device according to any one of claims 1 to 13. 15. The notifying means (150, 150a) and the alarm means (180, 200) are composed of a loudspeaker (4b) and an alarming device (3c) for outputting a voice, and the detecting means (3a). ) Is the microphone (3
By adding a second microphone (4d) for inputting a voice in the vicinity of the loudspeaker (4b), the communication means capable of bidirectional communication is provided between the inside and the outside of the detection space. The human body abnormality detection device according to claim 13 or 14, which is provided. 16. The loudspeaker (4) is a voice whose specific frequency component has been removed by a second filter means (5b) from the voice input by the microphone (3a).
16. The human body abnormality detection device according to claim 15, wherein the output is from b).