JP2002102187A - Living body detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reliably detect a living body even when output from a vibration sensor is low. SOLUTION: Time series data is converted into frequency series data with a frequency analyzing means 5 in order to evaluate the periodicity of the output from the vibration sensor 2 provided under a mattress 1, and featured values are extracted with a featured value extraction means 6 to detect the presence/ absence of a living body on the mattress 1 based on the amount of the featured value. By doing this, even when the mattress is placed on a hard tatami floor 10 and the vibration sensor 2 vibrates so hardly that the output is low, the presence/absence of the living body can be reliably detected by using the featured values periodically generated from breathing or heart beat activity.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a living body detecting device for detecting the presence / absence of a living body such as a seat and a bed.

[0002]

2. Description of the Related Art Conventionally, this kind of living body detecting apparatus has been disclosed, for example, in Japanese Patent Application Laid-Open No. 06-317676. FIG. 8 shows a block diagram of a conventional living body detection device. That is, reference numeral 20 denotes a flexible piezoelectric element disposed on the surface cloth 19 of the seat 18, reference numeral 21 denotes an amplifying means for amplifying an output signal of the piezoelectric element 20, and reference numeral 22 denotes a specific frequency component output from the amplifying means 21. 23 is a maximum displacement detecting means for detecting and outputting the maximum displacement of the output signal of the filter 22 for a certain period of time, and 24 is a human body which is a living body sitting on a seat from the output of the maximum displacement detecting means 23. It is a determination means for determining the presence or absence of.

FIG. 9 shows an output diagram of the maximum displacement detecting means 23 of the conventional living body detecting device. Here, when the human body is seated on the seat 18, the piezoelectric element 2 disposed on the surface cloth 19 of the seat 18.
0 is deformed, and a voltage is generated by the piezoelectric effect. In this generated signal, a large signal appears temporarily due to the impact of the sitting when the person is seated as shown in FIG. 9A, but when the human body is at rest, as shown by B, the heartbeat or respiration of the human body causes A fine body motion signal appears. If no human body is present, the output signal will be small and approach zero, as indicated by C. When an object is placed on the seat 10, a large signal appears temporarily as shown at D at the moment when the object is placed, but the object does not have minute movement due to heartbeat or breathing like a human body. So E
The output signal approaches zero again as shown in FIG. The output signal from the piezoelectric element 20 is amplified by the amplifying means 21 and filtered to a frequency component required by the filter 22,
The maximum displacement detecting means 23 detects the maximum displacement of the output of the filter 22, and using this value, the determining means 24 determines the presence or absence of a person.

[0004]

However, in the above-mentioned conventional living body detecting device, when the place where the piezoelectric element is laid is directly placed on a hard flooring such as a board or a tatami mat, the vibration of the vibration sensor is reduced. Because of the hindrance, there is a problem that even if the living body is on the biological signal detecting means, the output of the sensor becomes small and it becomes difficult to detect the biological signal. In particular, when trying to detect the presence or absence of a sleeping person laying on a bed, it is difficult to lay it on the human side of the bed because it is troublesome in terms of hygiene and daily raising and lowering. There has been a need for a system that can be provided with detection means.

[0005]

In order to solve the above-mentioned problems, the present invention has periodicity evaluation means for evaluating the magnitude of a periodically generated signal component of the biological signal detected by the biological signal detection means. The presence / absence of a living body on the supporting means is determined using at least one of the output of the periodicity evaluating means and the output of the biological signal detecting means.

According to the above invention, the presence / absence of a living body is determined by using the periodicity of a biological signal that periodically occurs such as the heartbeat and respiration of the living body.
Since the absence is determined, accurate determination can be performed even when the output signal of the biological signal detection unit is small as compared with the case where only the magnitude of the biological signal is used.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A living body detecting apparatus according to a first aspect of the present invention comprises a supporting means for supporting a living body, and a biological signal detecting means for detecting a biological signal on the supporting means due to respiration or heartbeat activity of the living body. A periodicity evaluation unit for evaluating the magnitude of a periodically generated signal among the biological signals detected by the biological signal detection unit, and at least one of an output of the periodicity evaluation unit and an output of the biological signal detection unit And determination means for determining the presence / absence of a living body on the support means using the output of Since the presence / absence of the living body is determined by using not only the size of the living body signal output means but also the periodicity of the living body signal which periodically occurs such as the heartbeat and respiration of the living body, only the size of the living body signal is used. As compared to the case, accurate determination can be made even when the output signal of the biological signal detecting means is small.

Further, the biological detecting apparatus according to a second aspect of the present invention is characterized in that the periodicity evaluating means analyzes the frequency of the biological signal detected by the biological signal detecting means, and the output of the frequency analyzing means. A determination means for determining the presence / absence of a living body on the support means; Then, since the periodicity of the biological signal detecting means is detected by extracting the characteristic amount of the output of the frequency analyzing means, it is possible to reliably evaluate the periodicity of the biological signal.

In the living body detecting apparatus according to a third aspect of the present invention, the feature extracting means has a sharpness calculating means for calculating the sharpness of the peak of the output of the frequency analyzing means. Since the biological signal exhibits strong periodicity, the sharpness of a sharp peak appearing on the frequency axis is used. Therefore, the characteristics of the biological signal can be easily extracted, and accurate biological detection can be performed.

In the living body detecting apparatus according to a fourth aspect of the present invention, the characteristic extracting means evaluates and extracts the periodicity of the output of the frequency analyzing means on the frequency axis. Since the periodicity of the biological signal is evaluated using the periodicity on the frequency axis, the periodicity of the biological signal can be more emphasized and extracted.

In the living body detecting apparatus according to a fifth aspect of the present invention, the characteristic extracting means calculates a frequency difference between the plurality of peaks and a peak extracting means for extracting a plurality of peaks of the output of the frequency analyzing means. A frequency difference calculating means, wherein the determining means uses the output of the peak extracting means or the frequency difference calculating means for determining the presence / absence of a living body on the supporting means. Since the frequency difference between a plurality of peaks is used, it is possible to extract the fundamental frequency component from several harmonic components appearing at a frequency that is a positive multiple of the fundamental frequency component, and the output of the fundamental frequency component is small. Even in such a case, the living body can be reliably detected. Further, even when the waveform of the fundamental frequency component due to respiration or heartbeat activity does not have a large peak, detection is possible if there are a plurality of peaks of the harmonic component, and it can be applied to variously changing biological signals. Further, there is no need to use an extremely low frequency signal such as breathing, so that processing can be performed in a short time, and there is an effect that a circuit configuration can be simplified such that a large capacitor is not required.

According to a sixth aspect of the present invention, in the living body detecting apparatus, the feature extracting means may be configured to perform the frequency analysis on the time series data with respect to the frequency series data output from the frequency analyzing means in the same manner as in the case of performing the frequency analysis. Is evaluated for periodicity. Then, by processing the frequency sequence data based on the fundamental frequency component and the harmonic component appearing at a positive multiple thereof in the same manner as the method of decomposing the time-series data into frequency components, the periodicity of the biological signal is more emphasized and extracted. It is possible to realize accurate living body detection. Further, even when the waveform of the fundamental frequency component due to respiration or heartbeat activity does not have a large peak, detection is possible if there are a plurality of peaks of the harmonic component, and it can be applied to variously changing biological signals. Further, there is no need to use an extremely low frequency signal such as breathing, so that processing can be performed in a short time, and there is an effect that a circuit configuration can be simplified such that a large capacitor is not required.

Further, the living body detection device according to claim 7 of the present invention calculates at least one of a heart rate or a heartbeat interval and a respiration rate or a breathing interval of the living body on the supporting means from the output of the feature extracting means. It has biological information calculation means. Then, by evaluating the periodicity of the biological signal,
At the same time as detecting the absence, it is possible to accurately measure the heart rate or heart rate interval or respiration rate or respiration interval of the living body from the periodicity.

Further, the living body detecting device according to claim 8 of the present invention has a large output detecting means for detecting that the output of the biological signal detecting means has exceeded a predetermined magnitude,
When the large output detecting means detects that the output of the living body detecting means has exceeded a predetermined magnitude, the determining means does not use the output of the periodicity evaluating means for determining the presence / absence of the living body on the supporting means. . Since the periodicity is not used for detecting a living body when it is difficult to obtain a biological signal having periodicity due to body motion, even if the periodicity of the output of the biological signal detecting means is disturbed and the periodicity cannot be detected, the living body can be reliably detected. Presence / absence can be detected.

Further, the living body detecting device according to claim 9 of the present invention has a large output detecting means for detecting that the output of the biological signal detecting means has exceeded a predetermined magnitude,
When the large output detecting means detects that the output of the living body detecting means has exceeded a predetermined magnitude, the biological information calculating means does not calculate the heart rate or the heart rate interval and the respiration rate or the breathing interval. In addition, when it is difficult to obtain a periodic biological signal due to body motion, the biological information such as heart rate and respiratory rate is not measured, so that the periodicity of the output of the biological signal detecting means is disturbed. No output.

According to a tenth aspect of the present invention, in the living body detection apparatus, the characteristic extracting means has a waveform extracting means for extracting a waveform in a predetermined frequency range from the output of the frequency analyzing means, Evaluate the periodicity of the output. Then, since the frequency range in which the characteristic is particularly strong is cut out from the output of the frequency analysis unit and the characteristic is extracted, the characteristic of the output waveform of the frequency analysis unit can be easily extracted.

[0017] In the living body detecting apparatus according to the eleventh aspect of the present invention, the waveform cutout means may be at least 0.1 Hz to 1 Hz.
Hz, more desirably, a waveform including a frequency range from 0.05 Hz to 2 Hz is cut out. The respiratory interval or respiratory frequency is calculated from the waveform of the output waveform of the frequency analysis means in which the periodic component due to respiration appears largely, so that the respiratory interval or respiratory frequency can be calculated accurately and easily.

According to a twelfth aspect of the present invention, in the living body detecting apparatus, the waveform cutout means may be at least 0.7 Hz to 3 Hz.
Hz, and more desirably, a waveform including a frequency range from 0.5 Hz to 8 Hz. Since the heartbeat interval or heart rate is calculated from the waveform of the output waveform of the frequency analysis means in which the periodic component due to the heartbeat activity appears largely, the heartbeat interval or heart rate can be calculated accurately and easily. And
Since a frequency band in which many peaks due to the heartbeat activity appear is extracted and its center frequency is largely extracted by a window function, a large output of the periodic component due to the heartbeat activity can be obtained.

In the living body detecting apparatus according to a thirteenth aspect of the present invention, the feature extracting means includes a window function calculating means for multiplying a waveform cut out by the waveform cutting means by a window function, and a cycle of an output of the window function calculating means. Assess gender and calculate respiratory rate or heart rate. Then, since the feature is extracted after emphasizing the portion where the feature of the respiratory or heartbeat activity appears strongly in the output of the waveform extracting means, the respiratory interval or respiratory rate, heartbeat interval or heart rate can be easily calculated.

[0020]

Embodiments of the present invention will be described below with reference to the drawings.

(Embodiment 1) FIG. 1 is a block diagram of a living body detecting apparatus according to Embodiment 1 of the present invention. In this embodiment, a case of a mattress is shown as a support means for supporting a living body. In the figure, reference numeral 1 denotes a mattress, which is a means for supporting a living body, 2 denotes a vibration sensor, which is a biological signal detecting means disposed below the mattress 1, 3
Is a signal processing means for amplifying the output signal of the vibration sensor and removing unnecessary frequency components, 4 is a amplitude calculating means for calculating the amplitude of the output of the signal processing means 3, and 5 is an output signal of the signal processing means 3. FF which is a periodicity evaluation means for converting into a digital value and evaluating the periodicity by fast Fourier transform
T means and 6 are feature extracting means for extracting the features of the FFT means 5, and 7 is a judging means for judging the presence / absence of a living body on the seat from the output of the swing width calculating means 4 and the output of the feature extracting means 6. , 8 is a biological information calculating means for calculating the respiratory rate and heart rate of the living body from the output of the feature extracting means, 9 is a display for displaying the determination result of the determining means 7 and the respiratory rate and heart rate calculated by the biological information calculating means 8. Means.

Here, the vibration sensor 2 is arranged below the mattress 1, and a tatami 10 is provided below the vibration sensor 2. The signal processing means 3 includes an amplifying means 3a for amplifying the output of the biological signal detecting means 2 and a filter 3b for removing unnecessary frequency signals. A sharpness detection means 6a for detecting a steep peak and calculating the sharpness of the peak by using the same, and a frequency difference calculation means 6b for calculating a frequency difference between adjacent peaks from an interval between a plurality of peaks in the frequency series. I have.

In this embodiment, the sharpness calculating means 6a also has a role of a peak extracting means for extracting a plurality of peaks of the output of the FFT means, and calculates the sharpness of the entire output waveform of the FFT means 5 to obtain the sharpness. A plurality of peaks are extracted by extracting a portion having a high peak.

The operation of the above configuration will be described. FIG. 2 shows an output diagram of the swing width calculating means, and FIG. 3 shows an output diagram of the FFT means. When a human body, which is a living body, lands on the mattress 1 and lays down or sits down, the vibration sensor 2 disposed below the mattress 1 vibrates to generate a voltage according to the magnitude of the vibration.
As shown in FIG. 2, a large signal appears temporarily in the generated signal due to the impact at the time of landing (A).
However, when the human body is at rest, a fine body motion signal due to the heartbeat and respiration of the human body appears (B). If there is no human body, the output signal becomes small and approaches zero (C).

On the other hand, when an object is placed on the mattress 1, a large signal appears temporarily at the moment when the object is placed (D). Since there is no motion, the output signal approaches zero again (E).

When a person walks on a mattress, it is the same as when an object is placed. The output signal from the vibration sensor 2 is amplified by the amplifying unit 3a of the signal processing unit 3, the unnecessary frequency components are removed by the filter 3b, and output to the amplitude calculating unit 4 and the FFT unit 5. The amplitude calculating means 4 calculates the amplitude of the output signal of the signal processing means and outputs it to the determining means 7.

On the other hand, in the FFT means 5, the output of the signal processing means 3 is subjected to fast Fourier transform at fixed time intervals to convert the time series data into frequency series data indicating the spectrum of the frequency component. Extract data features. The feature extraction by the feature extraction unit 6 includes a sharpness detection unit 6 that detects a steep peak from the frequency series data using the sharpness and calculates the sharpness of the peak.
a, and a peak interval detecting means 6b for detecting intervals between a plurality of peaks in the frequency sequence.

FIG. 3 (a) shows the output of the FFT means when a human body is present in the mattress 1. The horizontal axis in the figure is frequency (H
z), the vertical axis is the amplitude spectrum. Since the activity of the human body due to respiration and heartbeat occurs periodically and repeatedly, the output signal of the vibration sensor 2 indicates that if the human body is at rest without any vibration, only the heartbeat and the vibration of the breathing will be prominent. Even in the processing result obtained by performing the fast Fourier transform, the peak of the frequency corresponding to the cardiac cycle or the respiratory cycle forms a sharper peak than other frequencies. In FIG. 3A, A1 indicates a peak due to respiration, and B1 indicates a peak due to heartbeat activity.

In addition to the peaks of these fundamental frequencies, peaks due to harmonics that are integral multiples of the peaks are also represented by A2, A3, and B.
2, a sharp peak is formed similarly as shown by B3, but a plurality of peaks composed of these fundamental frequencies and their harmonics are equally spaced on the frequency axis at integer multiples of the fundamental frequency of breathing or heart activity. Appears in

On the other hand, FIG. 3B shows the output waveform of the FFT means when there is no human body on the mattress 1. In FIG. 3B, not only the output is small, but also the sharp peaks at equal intervals present in FIG. They do not exist at all, and they differ greatly in size and waveform. Even if the output of the biological signal detecting means is small, the characteristics of the waveform remain, so that the presence / absence of a living body can be reliably detected by examining the characteristics of such frequency sequence data and digitizing it. .

FIG. 4 shows a living body detection algorithm in the judgment means 7 of this embodiment. The power supply (not shown) of the device is O
When it becomes N and the apparatus starts operating (ST1), it is first recognized that the state is "tentatively present" (ST2), and the apparatus enters an absence determination routine for determining whether the apparatus is truly absent. Here, first, the absence determination using the output of the swing width calculating means 4 is performed (ST5). The output of the amplitude calculating means 4 is compared with a threshold V1 which is statistically determined in advance, and the feature extracting means is used only when the state in which the output of the amplitude calculating means 4 is equal to or less than V1 continues for a predetermined time T1 or more. The processing shifts to absence determination using the output of (1). In the absence determination using the output of the feature extracting means 6, first, the sharpness detecting means obtains the sharpness of each frequency data obtained by (Equation 1) from the frequency series of the output of the FFT means (ST7).

P _n = (f _n −f _na ) × | f _n −f _{n + a} | ² / (f _n −f _{n + a} ) (Equation 1) In Equation 1, P _n is the frequency n , F _n is FFT
The amplitude at the frequency n output by the means.
a is the width of the frequency serving as a reference for obtaining the sharpness, and in this embodiment, it is set to 0.2 Hz. In (Equation 1), the value becomes positive only when f _n is convex upward, and the greater the difference from the amplitude of the frequency component that becomes the base of the peak, the larger the value. If the sharpness is high, it is likely that periodic vibration due to breathing or heart rate activity is being applied,
There is a high possibility that the human body exists even if the swing width is small. The maximum value of the sharpness is compared with a predetermined threshold value P1. If the maximum value is lower than P1, it is determined that there is no possibility of the presence of a human body by this determination, and the process proceeds to the absence determination (ST9) based on the output of the peak interval detecting means. . In this determination, three peaks are extracted in ascending order of sharpness using the sharpness determined by the sharpness detecting means, and the differences df1 and df2 between the adjacent peaks among these peaks are calculated to calculate the frequency difference. Is being checked for a constant frequency or an integer multiple thereof.

Specifically, an index q obtained by the equation 2 is calculated, and when this value is larger than a predetermined threshold value Q1, it is determined that there is no possibility that a living body exists, and this determination is made. The absence is determined only when is performed (ST10). If the absence is determined from the “temporary existence” state (ST11), the provisional existence determination is discarded, and
The state is changed to absent determination retroactively to the point of time when "temporarily exists" (ST12).

Q = df1 / df2-int (df1 / df2) (Equation 2) The determination using the peak interval detecting means is performed even when the waveform of the fundamental frequency component due to respiration or heart activity does not have a large peak. If there are a plurality of peaks of the harmonic component, the detection is possible, and it can be widely applied to variously changing biological signals. Further, since it is not necessary to use an extremely low frequency signal such as breathing, processing can be performed in a short time, and there is an effect that a circuit configuration can be simplified and miniaturized such that a large capacitor is not required.

When the determination is absent, the process proceeds to the presence determination routine. In this determination, since a large vibration is applied when a person gets on the floor, a threshold value V2 is set in advance to detect this, and the output of the swing width calculation means is set to V
If the number exceeds 2 (ST13), the state is "temporarily a human body is present", and the process returns to the "absence determination routine" described above. However, in this case, if the process proceeds to ST10 within a certain period of time and the absence is determined, the large vibration detected in ST13 is recognized as a vibration other than a human body, and the presence of the large vibration is traced back to the provisional presence (ST12).

On the other hand, if the absence does not occur after a certain period of time, the presence determination is determined (ST14), and the person recognizes that the living body is present on the mattress. Once it is determined that "there is a person", a determination is made based on the output of the swing width calculating means (ST17), and when the output of the swing width detecting means falls below the threshold value V1 used in ST4, "temporarily absent". (ST18), the process proceeds to the first absence determination routine. Here, when the process proceeds to “absence determination” (ST10), the state changes to the absence determination. When the process returns to “absence determination” (ST14), the state of “temporarily absent” is discarded (ST15), and the status becomes “temporarily absent”. It returns to the presence determination retroactively to the time (ST
16).

By using the above algorithm, it is possible to reliably detect the presence of a living body on the mattress even when the biological signal is small. This is because if the mattress is on a hard tatami mat and the vibration sensor is sandwiched between the hard tatami mat and the mattress, the vibration due to the activity of living organisms will not be transmitted much to the vibration sensor, the output of the vibration sensor will also be small, and only the swing width will be In some cases, it was difficult to make a sufficient judgment. However, in the case of the present embodiment, the characteristics of the living body can be obtained even if the swing width is small, which is particularly effective. Such a determination result of the determination means 7 is output to the display means 9 and is visually displayed.

Further, in the living body detection device of this embodiment, when a living body is present on the mattress 1, the living body information calculating means 8 uses the output of the frequency difference calculating means 6b to calculate the respiration rate and heart rate of the living body on the mattress 1. The number of biological information is calculated. As described above, the output waveform of the FFT means 5 has a fundamental frequency due to respiration and heart rate activity and a plurality of peaks due to harmonics thereof, and the frequency difference between adjacent peaks corresponds to the reciprocal of the respiration rate or heart rate. Therefore, the frequency difference calculating means 6b
It is possible to easily calculate the respiratory rate and the heart rate by using the output of. The respiratory rate and heart rate calculated in this way are output to the display means 9 and visually displayed.

The judging means 7 has a large output detecting means 11 for detecting a case where the output of the swing width calculating means 4 exceeds a predetermined large output.
When detecting a large output, the absence determination by the evaluation of the periodicity and the calculation of the respiratory rate and the heart rate by the biological information calculating means 8 are stopped. This is because the periodicity of the output signal of the biological signal detecting means 2 by the FFT means 5 clearly appears only when the living body on the supporting means is at rest, and the living body has moved on the supporting means. In this case, the periodicity of the output of the biological signal detecting means is disturbed, so that correct absence determination and biological information calculation cannot be performed. Therefore, in such a case, the absence of presence determination is not performed and the conventional presence or absence determination is maintained, and the calculation is not performed due to the large output instead of outputting the respiratory rate and heart rate without performing the biological information calculation. Signal is being output. This realizes a living body detection device that does not output erroneous determinations or uncertain biological information even when there is a large output and the periodicity of the biological signal does not appear.

As described above, the living body detection device of this embodiment utilizes the periodicity of the biological signal periodically generated by the breathing of the living body and the activity of the heart, and the presence / absence of the living body on a supporting means such as a mattress. Is determined, the presence / absence of a living body can be reliably detected even when the output signal of the biological signal detecting means is small.

Also, the periodicity of the biological signal is evaluated by the sharpness detection means by quantifying the periodicity of the peak of the output waveform of the frequency analysis means, and the presence / absence of the living body is determined from the output of the biological signal detection means. It can be determined reliably. Further, even when the waveform of the fundamental frequency component due to respiration or heartbeat activity does not have a large peak, detection is possible if there are a plurality of peaks of the harmonic component, and it can be applied to variously changing biological signals. Further, there is no need to use an extremely low frequency signal such as breathing, so that processing can be performed in a short time, and there is an effect that a circuit configuration can be simplified such that a large capacitor is not required.

Further, the periodicity of the biological signal is quantified and evaluated by using the frequency difference between a plurality of peaks having high sharpness by the frequency difference calculating means, so that the presence or absence of the living body can be reliably determined from the output of the biological signal output means. Can be determined.

Further, the respiratory rate and the heart rate are calculated by the frequency difference calculating means using the frequency difference between a plurality of peaks having high sharpness. Can be calculated.

Further, since the case where periodicity cannot be evaluated from the output of the biological signal detecting means can be detected by the large output detecting means, it is possible to make an uncertain presence / absence judgment or to calculate an inaccurate respiration rate / heart rate. There is no.

In this embodiment, FF is used to evaluate the periodicity.
Although the evaluation is performed using the frequency analysis means by the T means, the evaluation can be performed using a method that does not perform conversion into a frequency sequence, such as an autocorrelation analysis method. Also in this case, the periodicity can be quantified and evaluated using the sharpness calculation and the interval between the peaks, and the respiration rate and the heart rate can be calculated. However, in this case, the peak difference is obtained in time.

In this embodiment, the FFT means is used as the frequency analysis means. However, a DFT (Discrete Fourier Transform) may be used, or the frequency analysis may be performed using an analog method using a plurality of band filters. May be performed.

Further, in this embodiment, the sharpness calculating means is F
Although the sharpness at all frequencies is obtained by applying the entire output waveform of the FT means, an algorithm that extracts a peak having a large swing width first and calculates the sharpness of the extracted peak may be used. The calculation formula of the sharpness is (Equation 1)
The formula is not limited to the above formula, and any formula can be used as long as the sharpness of the waveform can be quantified.

In this embodiment, the frequency difference calculating means calculates the frequency difference between adjacent peaks, but the frequency difference is not necessarily required to be adjacent peaks. However, in this case, all of the calculated frequency difference values may not coincide with the fundamental frequency. Therefore, it is desirable to increase the number of peaks for obtaining the fundamental frequency difference from three.

In this embodiment, a mattress is used as a means for supporting a living body, but it may be used for a bed or a seat. It is particularly effective when used for hard wooden beds and seats.

In this embodiment, the signal processing means comprises an amplifying means and a filter. However, the amplifying means is not always necessary when the sensitivity of the biological signal detecting means is good and sufficient output can be obtained, and a filter is not necessary. If the magnitude of the signal component is sufficiently smaller than the biological signal, it can be made unnecessary.

Further, in this embodiment, the biological information calculating means calculates both the respiratory rate and the heart rate of the living body on the supporting means, but it is also possible to calculate only one of the values.

In this embodiment, V is determined by the determination means.
Although presence / absence is determined using a plurality of predetermined thresholds such as 1, T1, etc., these thresholds are actually checked by a large number of people, and the optimal values statistically obtained from these data are used. are doing.

(Embodiment 2) Embodiment 2 of the present invention will be described with reference to the drawings.

The difference between this embodiment and the first embodiment is that
In order to extract the characteristics of the output waveform of the first FFT means 12 for fast Fourier transforming the output signal of the signal processing means, a waveform of approximately 0 to 1.5 Hz having a large characteristic due to respiration from the output waveform of the first FFT means Waveform extracting means 13 for extracting a waveform of approximately 0 to 8 Hz having a large characteristic due to the heartbeat, a first window function calculating means 14 for calculating a waveform of the respiratory component cut by the waveform extracting means 13 by applying a window function, To perform an FFT process on the output of the window function calculating means 14 of the second F
FT means 16, second window function calculating means 15 for applying a window function to the waveform of the heartbeat component cut out by waveform cutting means 13, and third processing for performing an FFT process on the output of second window function calculating means 15. The FFT means 17 is provided, and the judging means 7 comprises a swing width calculating means 4, a second FFT means 16 and a third FFT means.
The presence of a human body on the mattress 1 is determined using the output of the means 17, and the biological information calculation means calculates the respiratory interval from the output of the second FFT means 16 and the heartbeat interval from the third FFT means 17. It is in. The second FFT means and the third FFT means calculate the data of the frequency series by the same algorithm as the FFT, and the data obtained is a time-dimensional series.

As described in the first embodiment, when the output signal of the biological signal detecting means 2 is subjected to frequency analysis by the first FFT means 12, the peak of the fundamental frequency of respiratory or heartbeat activity and a frequency that is an integral multiple of the peak are obtained. The peak of the appearing harmonic component appears. Since peaks appear regularly at regular intervals in the frequency sequence waveforms of the fundamental frequency and harmonic components, when this frequency sequence data is further transformed by the same calculation method as the fast Fourier transform, the horizontal axis has a new time dimension. A waveform which is converted into a series and has an extremely large peak at a position corresponding to the interval of occurrence of respiration or heartbeat is obtained.

FIGS. 6A, 6B and 6C show the results of these conversions. FIG. 6A shows the time-series data of the output of the signal processing means 3, wherein the respiratory interval was 3.1 s on average and the heartbeat interval was 0.8 s on average. FIG. 6B shows FIG.
FIG. 6 shows frequency series data in which the absolute value of the result of fast Fourier transform of the waveform of FIG.
FIG. 6C shows the result of calculating the absolute value of the result of further fast Fourier transform of FIG. 6B using the FFT algorithm, and the horizontal axis is series data having a time dimension. FIG. 6 (b)
Although the fundamental frequency of respiration 0.32 Hz and its harmonics are clearly shown, the fundamental frequency of heart rate activity 1.25 Hz is not so clear, and the peak of the second harmonic of 2.5 Hz is large. . This is considered to be due to the fact that the natural frequency of the human body exists in the vicinity of 3 to 6 Hz, and a vibration component close to this frequency range is easily transmitted. This often occurs when the output of the biological signal detecting means 2 is small due to the difficulty in directly transmitting the vibration of the breathing or heartbeat activity to the vibration sensor 2 such as when lying down sideways.

However, FIG. 6 (c) showing the result of further fast Fourier transform of this frequency series data shows that
It can be seen that the largest peak exists near the heartbeat interval of 0.8 s, and a relatively large peak also exists near 3.1 s, which is close to the respiration interval. This is shown in FIG.
FIG. 6B shows that peaks are arranged at the same frequency interval as the fundamental frequency and the fundamental frequency due to the harmonic component as a waveform close to a sine wave whose cycle is the fundamental frequency.

The advantage of using this conversion is that even when the peak of the fundamental frequency is small or a part of the peaks are small, the periodicity on the frequency series can be numerically expressed as a whole.
The maximum peak of the obtained waveform is likely to appear at a position on the time axis corresponding to the interval between breathing and heartbeat. Also,
Since there is no need to extract peaks in advance, even when the peaks of the frequency sequence waveform after the fast Fourier transform are not so clear, it is possible to accurately extract the feature based on the fundamental frequency.

As described above, since the characteristic inherent to the living body due to respiration or heartbeat can be accurately obtained from the output waveform of the biological signal detecting means, the living body on the supporting means can be reliably detected.

Further, since the largest peak tends to appear at a position on the time axis corresponding to the interval between the occurrence of respiration and heartbeat, biological information of the interval between respiration and heartbeat can be easily calculated from the output waveform.

In this embodiment, before the FFT algorithm is directly applied to the output of the first FFT means, the frequency range of approximately 0 to 1.5 Hz and the characteristic of the heartbeat in which the characteristic of respiration appears in the waveform extracting means is strong. The two waveforms that appear only in the frequency range of approximately 0.5 to 8 Hz are cut out, and the first is
After multiplying the window function by the Hanning window by the window function calculating means and the second window function calculating means, the FFT algorithm is applied by the second FFT means and the third FFT means to convert the frequency sequence data into the time series data. Has been converted to.

This is because the fundamental frequency is different between the heartbeat and the respiration, and the frequency region where the harmonic component appears is also different. Therefore, a frequency region where the harmonic component is likely to appear between the heartbeat and the respiration is provided. This is because, when the output of the means is cut into a series in each frequency domain and then the Fourier transform is performed after applying the window function, it is easy to obtain the respiratory interval and the heartbeat interval individually. FIG. 7A shows the output of the first FFT means. FIG. 7B is a waveform obtained by cutting out a signal of approximately 0 to 1.5 Hz in which the fundamental frequency of respiration and its harmonic components are abundant from the frequency sequence data of FIG. 7A, and FIG. FIG. 7D shows a waveform obtained by cutting out a signal of about 0.5 to 8 Hz in which the fundamental frequency of the heartbeat activity and its harmonic components are abundant from the frequency series data of FIG. 7A, and FIG. 7 (e) is a waveform obtained by applying a Hanning window to FIG. 7 (c), FIG. 7 (f) is a waveform obtained by performing a fast Fourier transform on FIG. 7 (d), and FIG. This is a waveform obtained by performing fast Fourier transform on e). In a window function such as a Hanning window, the frequency near the center of the region is converted so as to be large, and the others are relatively small. Therefore, in FIG. Is gone. As a result, the peak frequency changes between FIG. 7 (f) and FIG. 7 (g).
In FIG. 7F, the largest peak exists at 3.1 s corresponding to the respiration interval, and in FIG. 7E, the largest peak exists at 0.8 s corresponding to the heartbeat interval, so that the respiration rate and the heart rate can be easily calculated. .

According to the above configuration, a frequency band in which many peaks due to respiration and heartbeat appear is extracted, and the center frequency thereof is largely extracted by a window function. Therefore, a large output of a periodic component due to respiration and heartbeat is obtained. In addition, it is possible to clarify the characteristics of the living body, and it is also possible to easily obtain the respiratory interval and the heartbeat interval.

In the above embodiment, the respiration interval and the heartbeat interval are obtained, but either one may be used. Further, the respiration rate and the heart rate, which are the reciprocals of the respiration interval and the heart rate interval, may be obtained.

Further, the first FFT means and the second FFT means may use a DFT or an analog method using a plurality of band filters. However, the second FF
Before realizing the T means, the output of the first FFT means needs to be continuous data of a frequency sequence, and it is not realistic to perform the second FFT means in an analog manner.

In this embodiment, the waveform extracting means is configured to extract two types of waveforms, one for respiration and the other for heartbeat, and to individually perform window function calculation and FFT processing. Or the configuration may be such that the processing of the respiration and the processing of the heartbeat are performed at different times with the same configuration.

The second and third FFT means can obtain substantially the same result even in the inverse Fourier transform except that the magnitude of the transform result is different. (Equation 3) shows the definition of DFT, and (Equation 4) shows the definition of inverse DFT. x (n) is a time series, X
[K] is a frequency sequence, and x (n) and X in (Equation 3)
When [k] is replaced, the only difference between these two expressions is the sign in parentheses of exp and 1 / N over the whole, and exp
The sign in parentheses can be ignored in the real part. FFT is DFT
Is a calculation method that speeds up the operation of, and is basically a calculation method that can obtain the same result.
Either T may be performed.

[0068]

(Equation 1)

[0069]

(Equation 2)

Further, in the above embodiment, when cutting out the data in the frequency domain, the frequency is set to 0 to 1.5 Hz in the case of breathing and 0.5 to 8 Hz in the case of heartbeat, but it is not limited to this frequency range. . For breathing, the fundamental frequency is 0.1 ~
Since it is about 1 Hz, the fundamental frequency is included in at least this range, so that it is possible to extract the periodicity due to respiration.

However, in order to calculate a more accurate value,
It is more advantageous to include the second harmonic, and the respiratory rate may be further reduced during sleep or the like. Therefore, it can be said that it is more preferable that the frequency range cut out for respiration includes 0.05 to 2 Hz.

The basic frequency of the heart rate is 0.7 to 3
Hz, the heart rate can be detected if this range is included at least, but the natural frequency of the human body is 4-6.
It is more computationally advantageous to include Hz, and it is more preferable to include 0.5 to 6 Hz in consideration of a decrease in heart rate during sleep. Further, a configuration may be employed in which the frequency range to be cut is changed according to the situation.

Further, in the above embodiment, the first and second window function calculating means each calculate using the Hanning window. However, the discontinuity of the waveform to be subjected to the FFT conversion such as the Hamming window or the Blackman window is eliminated. Another window function may be used as long as it is a window function that corrects and reduces the error of the FFT transform.

Further, in the above two embodiments, a vibration sensor, which is laid under the mattress and directly receives the vibration of the human body, is used as the biological signal detecting means. One that detects the human body by detecting the vertical movement of the abdomen of the human body, and one that detects the living body by detecting subtle changes caused by the human body's breathing and heartbeat activity on the futon in a non-contact manner using a ccd camera etc. The same effect can be obtained by applying.

As described in the above two embodiments, the living body detecting device of the present invention can surely detect a living body on a seat, a bed or a futon even when the output of the biological signal detecting means is small. By using this, it is possible to automatically perform thermal control or acoustic control in the presence / absence of a living body, or to search for a vacant seat in a vehicle or a movie theater. In addition, it is possible to check the presence time and occupancy time, especially the presence of the human body on a bed or futon.
If absent is determined, living conditions such as sleep can also be recorded. Furthermore, a system that uses the detected living conditions of the human body to notify a third person, such as a relative, of the situation when it recognizes that the living condition is different from daily life, such as a longer bedtime than usual, and warns the user to pay attention. It can also be applied to

[0076]

As explained in the above two embodiments, the living body detecting apparatus according to the first to thirteenth aspects of the present invention has the following effects.

The living body detecting device according to the first aspect of the present invention comprises:
Since the presence / absence of a living body is determined using not only the size of the biological signal output means but also the periodicity of a biological signal that periodically occurs, such as the heartbeat and respiration of the living body, the case where only the size of the biological signal is used In comparison, accurate determination can be made even when the output signal of the biological signal detecting means is small.

Further, the biological detection device according to claim 2 of the present invention extracts the characteristic amount of the output of the frequency analysis means and detects the periodicity of the biological signal detection means. Can be evaluated.

Further, the biological detection device according to claim 3 of the present invention uses the sharpness of a sharp peak appearing on the frequency axis because the biological signal exhibits strong periodicity, so that the characteristics of the biological signal can be easily extracted. And accurate living body detection can be performed.

Further, the biological detection device according to claim 4 of the present invention evaluates the periodicity of the biological signal using the periodicity on the frequency axis. Can be.

Further, the living body detecting apparatus according to claim 5 of the present invention uses a frequency difference between a plurality of peaks. The component can be extracted, and the living body can be reliably detected even when the output of the fundamental frequency component is small. Further, even when the waveform of the fundamental frequency component due to respiration or heartbeat activity does not have a large peak, detection is possible if there are a plurality of peaks of the harmonic component, and it can be applied to variously changing biological signals. Further, there is no need to use an extremely low frequency signal such as breathing, so that processing can be performed in a short time, and there is an effect that a circuit configuration can be simplified such that a large capacitor is not required.

Further, the living body detecting apparatus according to claim 6 of the present invention performs processing in the same manner as a method of decomposing frequency series data based on a fundamental frequency component and a harmonic component appearing as a positive multiple thereof from time series data into frequency components. By doing so, the periodicity of the biological signal can be more emphasized and extracted, and accurate biological detection can be realized. Further, even when the waveform of the fundamental frequency component due to respiration or heartbeat activity does not have a large peak, detection is possible if there are a plurality of peaks of the harmonic component, and it can be applied to variously changing biological signals. Further, there is no need to use an extremely low frequency signal such as breathing, so that processing can be performed in a short time, and there is an effect that a circuit configuration can be simplified such that a large capacitor is not required.

Further, the living body detecting apparatus according to claim 7 of the present invention detects the presence / absence of a living body by evaluating the periodicity of the living body signal, and at the same time, detects the heart rate or heartbeat interval of the living body and the respiratory rate from the periodicity. Alternatively, the breathing interval can be accurately measured.

Further, the living body detecting device according to the eighth aspect of the present invention does not use the periodicity for detecting a living body when it is difficult to obtain a periodic biological signal due to body movement or the like. However, the presence / absence of a living body can be reliably detected.

Further, the living body detecting apparatus according to the ninth aspect of the present invention does not measure the heart rate or the heart rate, the respiratory rate or the respiratory rate when it is difficult to obtain a periodic biological signal due to body movement or the like. It does not output uncertain values when it is difficult.

In the living body detecting apparatus according to the tenth aspect of the present invention, a frequency range where a characteristic is particularly strong is extracted from the output of the frequency analyzing means to extract the characteristic. Easy to extract.

In the living body detecting apparatus according to the eleventh aspect of the present invention, the respiratory interval or the respiratory frequency is calculated from the waveform of the output frequency of the call frequency analyzing means, in which the periodic component due to respiration appears largely. The interval or respiratory frequency can be calculated.

In the living body detecting apparatus according to the twelfth aspect of the present invention, the heartbeat interval or the heart rate is calculated from the waveform of the output waveform of the frequency analysis means, in which the periodic component due to the heartbeat activity appears largely. Interval or heart rate can be calculated.

In the living body detecting apparatus according to the thirteenth aspect of the present invention, since the features of the output of the waveform extracting means are emphasized at the portions where the characteristics of the respiratory or cardiac activity appear strongly, the respiratory interval can be easily determined. Alternatively, the respiration rate, heart rate interval or heart rate can be calculated.

[Brief description of the drawings]

FIG. 1 is a block diagram of a living body detection device according to a first embodiment of the present invention.

FIG. 2 is an output characteristic diagram of a swing width calculating unit of the apparatus.

FIG. 3A is an output characteristic diagram when the living body of the FFT unit of the apparatus is present; FIG.

FIG. 4 is a flowchart of a judgment means of the apparatus.

FIG. 5 is a block diagram of a living body detection device according to a second embodiment of the present invention.

6A is an output characteristic diagram of a signal processing unit of the device. FIG. 6B is an output characteristic diagram of a first FFT unit of the device. FIG. 6C is an output characteristic diagram of a second FFT unit of the device.

FIG. 7A is an output characteristic diagram of a first FFT unit of the device. FIG. 7B is an output characteristic diagram of a waveform extracting unit of the device for detecting respiration. FIG. 7C is a heartbeat detection of a waveform extracting unit of the device. (D) Output characteristic diagram of the first window function calculating means of the device (e) Output characteristic diagram of the second window function calculating means of the device (f) Calculation of the suction number of the device (F) Output characteristic diagram of second FFT means cut out for calculation (g) Second FFT cut out for heart rate calculation of the same device
Output characteristic diagram of means

FIG. 8 is a block diagram of a conventional living body detection device.

FIG. 9 is an output diagram of a signal processing unit of the apparatus.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 mattress (supporting means) 2 vibration sensor (biological signal detecting means) 5 FFT means (periodicity evaluating means) 6 feature extracting means 6a sharpness calculating means (peak extracting means) 6b frequency difference calculating means 7 determining means 8 biological information calculation Means 11 Large output detection means 12 First FFT means (periodicity evaluation means) 13 Waveform extraction means 14 First window function calculation means 15 Second window function calculation means 16 Second FFT means (feature extraction means) 17 Third FFT means (feature extraction means)

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Yumiko Hara 1006 Kazuma Kadoma, Kadoma-shi, Osaka Matsushita Electric Industrial Co., Ltd. F-term (reference) 4C017 AA10 AA14 AB10 AC03 BC16 BD06 FF05 4C038 VA16 VB31 4C040 AA17 AA30 GG20

Claims (13)

[Claims] 1. A supporting means for supporting a living body, a biological signal detecting means for detecting a biological signal due to a respiration or a heartbeat activity of the living body on the supporting means, and a periodic signal among the biological signals detected by the biological signal detecting means. A periodicity evaluating means for evaluating the magnitude of a signal generated in the living body, and presence / absence of a living body on the supporting means using at least one of an output of the periodicity evaluating means and an output of the biological signal detecting means. A living body detection device comprising: a determination unit that determines 2. The periodicity evaluation means comprises: frequency analysis means for frequency-analyzing the biological signal detected by the biological signal detection means;
2. The living body detection device according to claim 1, further comprising a feature extraction unit configured to extract a feature from an output of the frequency analysis unit, wherein the determination unit uses an output of the feature extraction unit to determine presence / absence of a living body on a support unit. . 3. The living body detection device according to claim 2, wherein the feature extracting means includes a sharpness calculating means for calculating the sharpness of the peak of the output of the frequency analyzing means. 4. The living body detection device according to claim 2, wherein the feature extraction unit evaluates and extracts periodicity of the output of the frequency analysis unit on a frequency axis. 5. The feature extracting means includes a peak extracting means for extracting a plurality of peaks of an output of the frequency analyzing means, and a frequency difference calculating means for calculating a frequency difference between the plurality of peaks. The living body detection device according to claim 4, wherein an output of the peak extracting means or the frequency difference calculating means is used for determining the presence / absence of a living body on the support means. 6. The living body according to claim 4, wherein the characteristic extracting means evaluates the periodicity on the frequency axis in the same manner as in the case where the frequency analysis of the time series data is performed on the frequency series data output from the frequency analyzing means. Detection device. 7. A biological information calculating means for calculating at least one of a heart rate or a heartbeat interval and a respiratory rate or a respiratory interval of the living body on the supporting means from an output of the feature extracting means. 2. The living body detection device according to claim 1. 8. A high-power detection means for detecting that an output of the biological signal detection means has exceeded a predetermined magnitude, wherein the large-output detection means has an output of the biological detection means having a predetermined magnitude. The living body detection device according to any one of claims 1 to 7, wherein when it is detected that the difference has exceeded the threshold value, the determination unit does not use the output of the periodicity evaluation unit to determine the presence / absence of a living body on the support unit. . 9. A high output detection means for detecting that an output of the biological signal detection means has exceeded a predetermined magnitude, wherein the high output detection means has an output of the biological detection means having a predetermined magnitude. The living body detection device according to claim 7, wherein the living body information calculation unit does not calculate the heart rate, the respiratory rate, and the like when detecting that the threshold value has been exceeded. 10. The feature extracting means has a waveform extracting means for extracting a waveform in a predetermined frequency range from the output of the frequency analyzing means, and evaluates the periodicity of the output of the waveform extracting means. 7. The living body detection device according to 5 or 6. 11. The waveform cutting means may have at least 0.1
Hz to 1Hz, more preferably 0.05Hz to 2Hz
The biological detection device according to claim 10, wherein a waveform including a frequency range up to is cut out. 12. The method according to claim 12, wherein the waveform cutting means is at least 0.7.
The living body detection device according to claim 10, wherein a waveform including a frequency range from Hz to 4Hz, more preferably from 0.5Hz to 8Hz is cut out. 13. The feature extracting means has a window function calculating means for multiplying a waveform cut out by the waveform cutting means by a window function, and extracts a feature of an output of the window function calculating means to extract a respiratory rate or a respiratory interval and a heart rate or heart rate. 13. The living body detection device according to claim 10, wherein at least one of a heartbeat interval is calculated.