KR20230133202A - Server, method and computer program for analyzing of sleep time using radar - Google Patents

Abstract

An apparatus for analyzing sleep time using radar includes a transmitter that transmits a radar signal toward a target object and receives a radar signal reflected from the target object; A breathing signal collection unit that collects a breathing signal for the target object based on the radar signal; a first sleep time detection unit that detects the first sleep time of the target object from a breathing signal based on the presence discriminator; a second sleep time detector that detects the second sleep time of the target object based on a probability value for the first sleep time; and a final sleep time determination unit that determines the final sleep time for the target object based on the second sleep time.

Description

Apparatus, method and computer program for analyzing sleep time using radar {SERVER, METHOD AND COMPUTER PROGRAM FOR ANALYZING OF SLEEP TIME USING RADAR}

The present invention relates to an apparatus, method, and computer program for analyzing sleep time using radar.

Polysomnography is a test that measures the quality and quantity of sleep and detects sleep diseases and sleep-related disorders. In general, various sleep diseases and sleep disorders are detected by measuring physiological and physical signals emitted from the human body during sleep. For example, brain waves, electrocardiogram, electromyogram, electrocardiogram, arterial blood, oxygen saturation, respiratory movements of the abdomen and chest, respiratory flow, snoring, and body posture are measured.

A basic method of measuring sleep time is to measure the wearer's sleep time based on the wearer's activity status (e.g., activity through wrist movement) while wearing a wearable device (e.g., wrist actigraphy, etc.). There is a way.

Conventional wearable-based sleep testing methods have the inconvenience of having to conduct the test with the testing device attached to a part of the human body, and have the limitation of relying only on expensive testing devices to more precisely test sleep diseases. exists.

Meanwhile, the radar-based sleep estimation method transmits a radar signal through a radar sensor, receives a radar signal reflected from a human object, and measures sleep time using the received radar signal.

This radar-based sleep estimation method does not constrain the human object because it does not require attaching a testing device to the human object, and the First Night Effect (FNE), which interferes with the sleep analysis of the human object, is very low. It is easy to analyze sleep.

However, because radar signals contain information about a wide area, there are false detection targets that can be confused with human objects.

Radar devices can filter stationary objects using signal processing, but cannot filter the movement of objects with subtle vibrations, such as the breathing of human objects. Here, the movement of objects with minute vibrations includes, for example, a curtain made of metal shaking in the wind, a plant with a high moisture content shaking in the wind, or a metal tong attached to a curtain that transmits radio waves being shaken by the wind. This may include shaking.

Briefly, referring to FIGS. 1A and 1B, these are diagrams to explain problems with radar signals used in existing sleep analysis.

Figure 1a is a diagram showing a radar signal pattern according to a plurality of setting environment patterns. Referring to FIG. 1A, (a) the remote activity pattern is a radar signal pattern for a sleep measurement subject measured in a nursing home shared room environment. In a nursing home shared room environment, the bed is only used for resting or sleeping, and other users may reside near the bed. In addition, it can be confirmed that when the sleep measurement subject resides in close proximity to the radar sensor, the sleep measurement subject's movement (i.e., the strength of the radar signal) is measured strongly, and at a long distance, the sleep measurement subject's movement is measured relatively weaker. You can.

(b) The side-lying pattern is a signal pattern that appears when a sleep measurement subject sleeps on his side in a single-person home environment. If you look at the side lying pattern, you can see that the breathing waveform of the sleep measurement subject is clearly observed, but the strength of the radar signal is measured very weakly.

(c) The false detection target pattern is the radar signal pattern for the false positive target measured in a nursing home multi-room environment. False detection target patterns include no breathing patterns for long periods of time and irregular noise patterns. This false detection target pattern is, for example, a pattern that can be used to estimate that an object presumed to be a large-area curtain or small metal tongs is continuously shaking due to the influence of wind around it.

(d) The stable standard pattern is a signal pattern that appears when there are no false positives around the bed in a single-person home environment and the sleep measurement subject sleeps while maintaining an upright posture (supine) while sleeping. In the case of a stable standard pattern, it can be seen that the movement and stationary state of the sleep measurement subject and the intensity of noise when away are clearly distinguished.

FIG. 1B shows the signal strength of the Normal Presence State, the signal strength of the Absence State, and the signal strength of the Movement State measured from each radar signal for each of the plurality of setting environment patterns in FIG. 1A. This is a drawing showing .

Referring to Figure 1b, the signal strength 101 of the absent state in the false detection target pattern is the signal strength 103 of the normal presence state in the side lying pattern and the distant activity pattern. It is greater than the signal strength (105) of the movement state in . This means that a person's occupancy status cannot be confirmed simply by comparing the intensity of the measured signal.

Korean Patent Publication No. 10-2014-0087902 (published on July 9, 2014)

The present invention is intended to solve the problems of the prior art described above, and collects breathing signals for a target object related to sleep using a radar, and sleep time for the target object from the breathing signal for the target object using an occupancy discriminator. I would like to analyze.

However, the technical challenges that this embodiment aims to achieve are not limited to the technical challenges described above, and other technical challenges may exist.

As a technical means for achieving the above-described technical problem, the device for analyzing sleep time using radar according to the first aspect of the present invention transmits a radar signal toward a target object and receives a radar signal reflected from the target object. a receiving sender; A breathing signal collection unit that collects a breathing signal for the target object based on the radar signal; a first sleep time detection unit that detects a first sleep time of the target object from the breathing signal based on an presence discriminator; a second sleep time detector that detects a second sleep time of the target object based on a probability value for the first sleep time; and a final sleep time determination unit that determines the final sleep time for the target object based on the second sleep time.

A method of analyzing sleep time using radar performed by a sleep time analysis device according to a second aspect of the present invention includes the steps of transmitting a radar signal toward a target object and receiving a radar signal reflected from the target object; Collecting a breathing signal for the target object based on the radar signal; detecting the first sleep time of the target object from the breathing signal based on an presence discriminator; detecting a second sleep time of the target object based on a probability value for the first sleep time; and determining the final sleep time for the target object based on the second sleep time.

A computer program stored in a computer-readable recording medium including a sequence of instructions for analyzing sleep time using radar according to the third aspect of the present invention, when executed by a computing device, transmits a radar signal toward a target object, Receive a radar signal reflected from the target object, collect a breathing signal for the target object based on the radar signal, and detect the first sleep time of the target object from the breathing signal based on an presence discriminator. A sequence of instructions for detecting the second sleep time of the target object based on the probability value for the first sleep time and determining the final sleep time for the target object based on the second sleep time. may include.

The above-described means for solving the problem are merely illustrative and should not be construed as limiting the present invention. In addition to the exemplary embodiments described above, there may be additional embodiments described in the drawings and detailed description of the invention.

According to one of the means for solving the problems of the present invention described above, the present invention collects breathing signals for a target object related to sleep using a radar, and uses a presence discriminator to determine the target object from the breathing signal for the target object. Sleep time can be analyzed.

In addition, the present invention can determine whether the target object is occupied by using an occupancy discriminator in cases where the occupancy state of the target object cannot be determined only by the strength of the existing radar signal.

In addition, the present invention can easily collect the sleep breathing of a target object in daily life using radar, so the sleep time of the target object can be easily analyzed non-face-to-face.

In addition, the present invention uses radar to analyze breathing signals related to the target object's sleep, thereby simplifying the inconvenience of conducting a conventional polysomnographic examination.

1A to 1B are diagrams to explain problems with radar signals used in existing sleep analysis.
2A to 2D are diagrams for explaining a method of calculating a normalized discriminator to be used in an presence discriminator according to an embodiment of the present invention.
Figure 3 is a block diagram of a sleep time analysis device according to an embodiment of the present invention.
Figures 4A to 4D are diagrams for explaining a method of determining the final sleep time according to an embodiment of the present invention.
Figure 5 is a flowchart showing a method of analyzing sleep time using radar, according to an embodiment of the present invention.

Below, with reference to the attached drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily implement the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. In order to clearly explain the present invention in the drawings, parts that are not related to the description are omitted, and similar parts are given similar reference numerals throughout the specification.

Throughout the specification, when a part is said to be "connected" to another part, this includes not only cases where it is "directly connected," but also cases where it is "electrically connected" with another element in between. . Additionally, when a part "includes" a certain component, this means that it may further include other components rather than excluding other components, unless specifically stated to the contrary.

In this specification, 'part' includes a unit realized by hardware, a unit realized by software, and a unit realized using both. Additionally, one unit may be realized using two or more pieces of hardware, and two or more units may be realized using one piece of hardware.

In this specification, some of the operations or functions described as being performed by a terminal or device may instead be performed on a server connected to the terminal or device. Likewise, some of the operations or functions described as being performed by the server may also be performed on a terminal or device connected to the server.

Hereinafter, specific details for implementing the present invention will be described with reference to the attached configuration diagram or processing flow diagram.

2A to 2D are diagrams for explaining a method of calculating a normalized discriminator to be used in an presence discriminator according to an embodiment of the present invention.

Figure 2a shows the radar signal 201-1 in the normal presence state, the radar signal 203-1 in the away state, and the movement state (201-1) for each setting environment pattern in Figures 1a and 1b. This is an enlarged version of the radar signal (205-1) of Movement State and is a graph of the breathing waveform of the sleep measurement subject. Referring to FIG. 2A, the radar signal 201-1 in a normal occupancy state for each setting environment pattern is a signal measured in a stable occupancy state in which the user is resting or sleeping.

Looking at the breathing waveform of the side-lying pattern 207 among the radar signals 201-1 in normal occupancy conditions, the radar signal is measured while the sleep measurement subject is lying on his side, so the size of the signal is small.

Looking at the breathing waveform of the false detection target pattern 209 among the radar signals 201-1 in normal occupancy conditions, it includes a section where apnea occurred during the sleep measurement subject's sleep.

In this way, it can be confirmed that the waveform pattern formed by breathing tends to be measured more clearly when the sleep measurement subject is in a stable state.

On the contrary, the wave pattern formed by noise or movement has an irregular pattern.

Figure 2b is a graph showing the results of filtering the radar signal 201-1 in the normal occupancy state, the radar signal 203-1 in the away state, and the radar signal 205-1 in the motion state for each setting environment pattern in Figure 2a. (201-3, 203-3, 205-3). Referring to FIGS. 2A and 2B, the sleep time analysis device of the present invention uses radar signals (201-1, 203-1, and 205-1) measured for each set environment pattern to clearly process the periodicity of the breathing band. Filtering can be done through [Equation 1].

[Equation 1]

At this time, the sleep time analysis device converts the radar signals (201-1, 203-1, 205-1) in decibel units to ADC units to further expand the change in amplitude, and then applies a band-pass filter. Filtered radar signals (201-3, 203-3, 205-3) can be derived. Here, the band-pass filter is, for example, a 21st order Finite Impulse Response (FIR) band-pass filter, and is in the 10 to 30 bpm frequency band, which is a representative breathing band (i.e., in the range of the typical respiratory rate, 10 times per minute). Only the band containing the signal between 30 times is left and the remaining frequency bands are removed.

The sleep time analysis device differentiates the filtered radar signals (201-3, 203-3, 205-3) for each set environment pattern to remove the direct current component, and converts the signal from which the direct current component has been removed into a signal that oscillates based on 0. It can be converted to . This can prevent the calculated value from being meaninglessly increased due to the direct current (DC) component.

The sleep time analysis device uses autocorrelation, which means that the autocorrelation value changes as the signal has a clearer periodicity in the same amplitude range. Here, autocorrelation, also called serial correlation, is a type of convolution calculated by crossing the delayed copy signal as a time delay function.

The signal differentiated from the filtered radar signals 201-3, 203-3, and 205-3 can be expressed by an equation such as [Equation 2].

[Equation 2]

[Equation 2] is the concept of differentiation, and in the discrete domain, = 1, and if this is real-time, it can be derived from an equation such as [Equation 3].

[Equation 3]

Sleep time analysis device is a differential signal The autocorrelation signal can be calculated using . Here, the autocorrelation signal can be expressed as [Equation 4].

[Equation 4]

FIG. 2C is a graph showing the autocorrelation signal 201-5 in the normal presence state, the autocorrelation signal 203-5 in the away state, and the autocorrelation signal 205-5 in the movement state for each setting environment pattern.

The clearer the periodicity of an autocorrelation signal, the more even the width of each peak and valley, and the width is narrower than that of signals with less clear periodicity.

To reflect this phenomenon, the sleep time analysis device may calculate the periodicity factor by calculating the sum of absolute values for the autocorrelation signal (including both positive and negative direction components). Here, the periodicity factor is a factor indicating the degree of periodicity in the same amplitude range and can be expressed as [Equation 5].

[Equation 5]

The sleep time analysis device can calculate a squared sum for the autocorrelation signal and calculate a normalization factor using the calculated squared sum. Here, the normalization factor is a factor for normalizing amplitudes of different ranges and can be expressed as [Equation 6].

[Equation 6]

Periodicity factor ( ) and normalization factor ( ) can be schematized as shown in the first graph 211 of FIG. 2D.

To calculate the general degree of periodicity, the sleep time analyzer divides the periodicity factor by the normalization factor to determine the normalized discriminant ( ) can be calculated. Here, the normalized discriminant can be expressed as [Equation 7].

[Equation 7]

The final derived normalized discriminant can be schematized as shown in the second graph 213 of FIG. 2D.

The present invention can distinguish cases in which a person's occupancy status cannot be determined only by the strength of the existing signal (FIG. 1b) through a normalized discriminator.

Looking at the second graph 213 of FIG. 2D, it can be quantitatively confirmed that the periodicity is more evident in the cases of stable occupancy shown in green.

Hereinafter, in order to estimate the sleep time of the target object, the present invention equates the state of no movement of the target object with a stable occupancy state and detects the stable occupancy state using the normalized discriminator described above.

Figure 3 is a block diagram of a sleep time analysis device 30 according to an embodiment of the present invention.

Referring to FIG. 3, the sleep time analysis device 30 includes a transceiver 300, a respiration collection unit 310, an occupancy discriminator derivation unit 320, a first sleep time detection unit 330, and a windowing likelihood. ) It may include a calculation unit 340, a second sleep time detection unit 350, a sleep continuity analysis unit 360, and a final sleep time determination unit 370. However, the sleep time analysis device 30 shown in FIG. 3 is only one implementation example of the present invention, and various modifications are possible based on the components shown in FIG. 3.

Hereinafter, FIG. 3 will be described with reference to FIGS. 4A to 4D.

The transceiver unit 300 may transmit a radar signal toward a target object and receive a radar signal reflected from the target object. For example, the transceiver 300 may transmit a radar signal toward a target object using a radar and receive a radar signal reflected from the target object through the radar.

Referring to FIG. 4A, the respiration collection unit 310 may collect the respiration signal 401 for the target object based on the radar signal. For example, the respiration collection unit 310 may collect respiration signals for 24 hours based on radar signals received from noon of the previous day to noon of the same day.

The occupancy discriminator deriving unit 320 may apply the concept of moving window settings when calculating the occupancy discriminator from the breathing signal 401.

Referring to FIGS. 4A and 4B together, the presence discriminator deriving unit 320 divides the respiration signal 401 into a plurality of segmented respiration signals based on the window size 40 corresponding to the preset first time unit, , Among the plurality of segmented breathing signals, two or more segmented breathing signals adjacent to each other may be arranged to overlap at a preset second time unit (42, overlap size) interval. Here, the first time unit may include 2 minutes, and the second time unit may include 1 minute.

The occupancy discriminator deriving unit 320 extracts the occupancy discriminator signal 403 of the second time unit from two or more overlapped segmented breathing signals, and determines occupancy from the extracted occupancy discriminator signal 403 of the second time unit. 405 can be derived. Here, the occupancy discriminator 405 has the characteristic of numerically separating signals in a stable occupancy state, signals in an occupancy state with movement, and signals in an away state with a single threshold for determining an occupancy state without movement. .

The first sleep time detector 330 may detect the first sleep time 407 of the target object from the breathing signal 401 based on the presence discriminator 405.

The moving likelihood calculation unit 340 may infer a probability value for the first sleep time 407 based on the preset third time unit. Here, the third time unit may include 30 minutes.

Referring to FIGS. 4A and 4C together, the second sleep time detector 350 may detect the second sleep time 413 of the target object based on the probability value for the first sleep time 407.

The windowing likelihood calculation unit 340 may calculate a windowing likelihood 409 indicating the sleep state from the probability value for the first sleep time 407. Here, the windowing likelihood 409 can be used to probabilistically estimate the sleep state to the degree that a stable occupancy state is dominant.

Looking at the graph of Windowing Likehood 409, the higher the likelihood value (i.e., maintaining a stable occupancy state for 30 minutes), the higher the likelihood of being in a sleeping state.

However, if the target object briefly uses the bed to rest, there are cases where the calculation is similar to a sleeping state, and if the target object moves while sleeping (especially lap sleep), it is difficult to judge it as continuous sleep. .

Therefore, the present invention uses the likelihood value for a stable presence state to determine continuous sleep when the movement or absence of the target object is temporary for 30 minutes.

Meanwhile, although the temporary state change of the target object was judged to be included in continuous sleep, a situation may actually occur in which the target object may go to the bathroom or move for more than a certain period of time due to external stimulation during sleep. In this situation, the present invention uses a double threshold technique to minimize confusion in the presence discriminator. Here, the double threshold technique is a technique used to reduce the phenomenon of frequent changes in results detected near a single threshold due to noise when detecting a threshold from a signal.

The second sleep time detector 350 may detect the second sleep time 413 of the target object from the moving likelihood 409 using a preset double threshold 411. Here, the double threshold 411 is used to determine the sleep state according to how long the temporary stable state is maintained by reducing the fluctuating situation. The double threshold 411 includes a first time threshold (Upper Threshold) and a second time threshold (Lower Threshold). Threshold) may be included. The first time threshold may include, for example, 7 minutes, and the second time threshold may include, for example, 12 minutes.

The second sleep time detection unit 350 determines it as an awakening state time when the sleep time corresponding to the windowing likelihood 409 is less than the first time threshold for the third time unit, and responds to the windowing likelihood 409. If the sleep time exceeds the second time threshold for the third time unit, it may be determined as the sleep state time.

For example, the second sleep time detection unit 350 may determine the sleep state as a sleep state if the time determined to be in a stable occupancy state is more than 12 minutes over 30 minutes, and may determine it as an awakening time if it is less than 7 minutes during 30 minutes. .

In this way, the present invention applies the double threshold 411 based on the windowing likelihood 409, converting the first sleep time 407 of the target object into a second sleep time having a density-based continuous form. It can be converted to sleep time (413). This is to correct errors in the occupancy discriminator and temporary situation changes.

Meanwhile, sleep patterns vary from person to person. A personally satisfactory sleep time (i.e., a sleep time that does not cause problems for an individual's daily life) and the intermediate awakening time to end sleep are different for each individual.

For example, insomnia may be diagnosed based on the degree of discomfort in daily life due to sleep rather than the objectively measured sleep time. Based on this phenomenon, it is necessary to consider the number of various cases of sleep time personally recognized by the target object regarding sleep continuity.

To this end, the sleep continuity analysis unit 360 can analyze various sleep patterns through mathematical morphology analysis. Mathematical morphology techniques include Dilation and Erosion operations, and Closing: Dilation → Erosion and Opening: Erosion → Dilation, which are a combination of the two operations.

In the mathematical morphology technique, boozing and melting operations are applied both before and after the detection area, so when implementing the actual function, only half of the interval to be removed is used.

Referring to FIG. 4C, the sleep continuity analysis unit 360 may analyze sleep continuity from the second sleep time 413 by considering a plurality of preset awakening integration times.

The sleep continuity analysis unit 360 can analyze sleep continuity by applying the first awakening integration time and the second awakening integration time in the second sleep time 413 using a closing operation, which is a mathematical morphology technique. . Here, the first awakening integration time may include 30 minutes, and the second awakening integration time may include 60 minutes.

If the awakening time included in the second sleep time 413 is less than one of the plurality of awakening integration times, the sleep continuity analysis unit 360 may process the awakening time to be integrated into a continuous sleep time.

For example, the sleep continuity analysis unit 360 may determine the basic inferred sleep time 417, which is a sleep time lasting more than 5 hours from the second sleep time 413, as the first candidate sleep time. At this time, the default inferred sleep time (417) is 03:59 [21:20-01:18].

The sleep continuity analysis unit 360 integrates awakening times of less than 30 minutes among the awakening times included in the second sleep time 413 into continuous sleep time, and integrates 30-minute awakenings when the continuous sleep time is 5 hours or more. Time 419 can be determined as the second candidate sleep time. At this time, the 30-minute awakening integration time (419) is 07:51 [21:20-05:09].

In addition, the sleep continuity analysis unit 360 integrates the awakening time of less than 60 minutes among the awakening times included in the second sleep time 413 into continuous sleep time, and if the continuous sleep time is 5 hours or more, 60 minutes. The minute awakening integration time 421 can be determined as the third candidate sleep time. At this time, the 60-minute awakening integration time (421) is 15:05 [17:42-08:46].

When determining sleep time, the sleep continuity analysis unit 360 ignores sleep time of less than 3 hours, does not recognize sleep segmentation of more than 1 hour as continuous sleep, and detects continuous sleep of more than 5 hours as candidate sleep. It can be confirmed by time.

This method of determining sleep time can be schematized as shown in FIG. 4D.

Time setting parameters, such as 30 minutes and 60 minutes corresponding to the sleep segment integration time described in the example above, and 3 hours and 5 hours corresponding to the minimum sleep time, can be continuously updated in the actual execution environment. Here, the time setting parameter may be adjusted, for example, according to feedback from the target object, may be adjusted according to a setting value directly by the target object, or may be a value derived from data of another target object with high similarity to the target object. It can be adjusted based on, and can be adjusted with representative parameters derived from a group using individual variables (e.g., gender, age, weight, etc.), and survey information of the target object (e.g., disease information of the target object, It can also be adjusted with representative parameters derived from similar groups based on usual sleep perception, insomnia, etc.).

The final sleep time determination unit 370 may determine the final sleep time for the target object based on the second sleep time 413.

The final sleep time determination unit 370 may determine the final sleep time further based on the analysis result of sleep continuity.

Referring to FIG. 4C, the sleep continuity analysis result 415 includes a basic inferred sleep time 417 corresponding to the first candidate sleep time, a 30-minute awakening integration time 419 corresponding to the second candidate sleep time, and A 60-minute wake integration time (421) corresponding to a third candidate sleep time may be included.

The final sleep time determination unit 370 may determine the longest sleep time among the first to third candidate sleep times 417, 419, and 421 as the final sleep time.

Meanwhile, those skilled in the art will know that the transceiver unit 300, the respiration collection unit 310, the presence discriminator derivation unit 320, the first sleep time detection unit 330, the moving likelihood calculation unit 340, the second It will be fully understood that the sleep time detection unit 350, the sleep continuity analysis unit 360, and the final sleep time determination unit 370 may be implemented separately, or one or more of them may be integrated and implemented.

Figure 5 is a flowchart showing a method of analyzing sleep time using radar, according to an embodiment of the present invention.

Referring to FIG. 5 , in step S501, the sleep time analysis device 30 may transmit a radar signal toward a target object and receive a radar signal reflected from the target object.

In step S503, the sleep time analysis device 30 may collect a breathing signal for the target object based on the radar signal.

In step S505, the sleep time analysis device 30 may detect the first sleep time of the target object from the breathing signal based on the presence discriminator.

In step S507, the sleep time analysis device 30 may detect the second sleep time of the target object based on the probability value for the first sleep time.

In step S509, the sleep time analysis device 30 may determine the final sleep time for the target object based on the second sleep time.

In the above description, steps S501 to S509 may be further divided into additional steps or combined into fewer steps, depending on the implementation of the present invention. Additionally, some steps may be omitted or the order between steps may be changed as needed.

One embodiment of the present invention may also be implemented in the form of a recording medium containing instructions executable by a computer, such as program modules executed by a computer. Computer-readable media can be any available media that can be accessed by a computer and includes both volatile and non-volatile media, removable and non-removable media. Additionally, computer-readable media may include all computer storage media. Computer storage media includes both volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data.

The description of the present invention described above is for illustrative purposes, and those skilled in the art will understand that the present invention can be easily modified into other specific forms without changing the technical idea or essential features of the present invention. will be. Therefore, the embodiments described above should be understood in all respects as illustrative and not restrictive. For example, each component described as unitary may be implemented in a distributed manner, and similarly, components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the claims described later rather than the detailed description, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be construed as being included in the scope of the present invention. .

30: Sleep time analysis device
300: Transmitter and receiver
310: respiratory collection unit
320: Occupancy discriminator derivation unit
330: First sleep time detection unit
340: Movement likelihood calculation unit
350: Second sleep time detection unit
360: Sleep continuity analysis unit
370: Final sleep time determination unit

Claims (21)

In a device that analyzes sleep time using radar,
a transmitter that transmits a radar signal toward a target object and receives a radar signal reflected from the target object;
A breathing signal collection unit that collects a breathing signal for the target object based on the radar signal;
a first sleep time detection unit that detects a first sleep time of the target object from the breathing signal based on an presence discriminator;
a second sleep time detector that detects a second sleep time of the target object based on a probability value for the first sleep time; and
A final sleep time determination unit that determines the final sleep time for the target object based on the second sleep time.
A sleep time analysis device comprising a.
According to claim 1,
Splitting the breathing signal into a plurality of segmented breathing signals based on a window size corresponding to a preset first time unit, and dividing two or more segmented breathing signals adjacent to each other among the plurality of segmented breathing signals at preset second time unit intervals It further includes an occupancy discriminator derivation unit arranged by overlapping,
The first time unit includes 2 minutes, and the second time unit includes 1 minute.
According to claim 2,
The occupancy discriminator deriving unit is configured to derive the occupancy discriminator from the overlapped two or more segmented breathing signals.
According to claim 1,
The sleep time analysis device further includes a movement likelihood calculation unit that calculates a movement likelihood indicating a sleep state from the probability value for the first sleep time.
According to claim 4,
The movement likelihood calculation unit infers a probability value for the first sleep time based on a preset third time unit,
Sleep time analysis device, wherein the third time unit includes 30 minutes.
According to claim 4,
The second sleep time detection unit detects the second sleep time of the target object from the movement likelihood using a preset double threshold.
According to claim 6,
the dual threshold includes a first time threshold and a second time threshold,
The second sleep time detector determines the waking state time when the sleep time corresponding to the movement likelihood is less than the first time threshold during the third time unit, and the sleep time corresponding to the movement likelihood is less than the first time threshold during the third time unit. A sleep time analysis device that determines sleep state time when the second time threshold is exceeded.
According to claim 1,
A sleep time analysis device further comprising a sleep continuity analysis unit that analyzes sleep continuity from the second sleep time in consideration of a plurality of preset awakening integration times.
According to claim 8,
The sleep continuity analysis unit is configured to process the awakening time included in the second sleep time to integrate the awakening time into a continuous sleep time when it is less than one of the plurality of awakening integration times.
According to claim 8,
The final sleep time determination unit determines the final sleep time based on the analysis result of the sleep continuity.
In a method of analyzing sleep time using radar performed by a sleep time analysis device,
Transmitting a radar signal toward a target object and receiving a radar signal reflected from the target object;
Collecting a breathing signal for the target object based on the radar signal;
detecting the first sleep time of the target object from the breathing signal based on an presence discriminator;
detecting a second sleep time of the target object based on a probability value for the first sleep time; and
Determining the final sleep time for the target object based on the second sleep time
Sleep time analysis method including.
According to claim 11,
Splitting the breathing signal into a plurality of segmented breathing signals based on a window size corresponding to a preset first time unit, and dividing two or more segmented breathing signals adjacent to each other among the plurality of segmented breathing signals at preset second time unit intervals Further comprising the step of overlapping and placing,
The first time unit includes 2 minutes, and the second time unit includes 1 minute.
According to claim 12,
A sleep time analysis method further comprising deriving the presence discriminator from the two or more overlapping segmented breathing signals.
According to claim 11,
Further comprising the step of inferring a probability value for the first sleep time based on a preset third time unit,
A sleep time analysis method, wherein the third time unit includes 30 minutes.
According to claim 14,
A sleep time analysis method further comprising calculating a windowing likelihood indicating a sleep state from the probability value for the first sleep time.
According to claim 15,
The step of detecting the second sleep time of the target object is
A sleep time analysis method comprising detecting the secondary sleep time of the target object from the movement likelihood using a preset double threshold.
According to claim 16,
the dual threshold includes a first time threshold and a second time threshold,
determining a wake state time if the sleep time corresponding to the movement likelihood is less than a first time threshold during the third time unit;
The sleep time analysis method further includes determining a sleep state time when the sleep time corresponding to the movement likelihood exceeds a second time threshold during the third time unit.
According to claim 11,
A sleep time analysis method further comprising analyzing sleep continuity from the second sleep time in consideration of a plurality of preset awakening integration times.
According to claim 18,
If the awakening time included in the secondary sleep time is less than one of the plurality of awakening integration times, processing the waking time to integrate into continuous sleep time.
According to claim 18,
The step of determining the final sleep time for the target object is
A sleep time analysis method comprising determining the final sleep time further based on the analysis result of the sleep continuity.
A computer program stored in a computer-readable recording medium containing a sequence of instructions for analyzing sleep time using radar,
When the computer program is executed by a computing device,
Transmitting a radar signal toward a target object and receiving a radar signal reflected from the target object,
Collecting respiration signals for the target object based on the radar signal,
Detecting the first sleep time of the target object from the breathing signal based on the presence discriminator,
Detecting the second sleep time of the target object based on the probability value for the first sleep time,
A computer program stored in a computer-readable recording medium, comprising a sequence of instructions for determining a final sleep time for the target object based on the second sleep time.