Fall detecting device and fall detecting method

Abstract

Provided is a fall detecting device and a fall detecting method having a reduced computational load. The fall detecting device comprises: a sensor; a signal strength calculating unit that calculates a signal strength for each of a plurality of range bins within a detection target area on the basis of transmission and reception signals of the sensor; a signal strength variation amount calculating unit that calculates a signal strength variation amount for each range bin; an area setting unit that sets a first area, within the detection target area, that does not include a range bin in which the signal strength is equal to or greater than a signal strength threshold; and a determining unit that detects the occurrence of a fall of a living body within the detection target area on the basis of a first signal strength variation amount, which is a statistical value of the signal strength variation amount for each of the plurality of range bins within the first area.

Description

Fall detection device and fall detection method

[0001] The present invention relates to a fall detection device and a fall detection method.

[0002] Accidents and falls in private rooms are a problem in various settings. For example, in nursing homes, residents spend a lot of time in their private rooms. Elderly residents and those requiring care are particularly susceptible to falls due to declining physical and cognitive functions. In factories, workers often work alone in their private rooms for long periods of time, posing a risk of falls and unexpected accidents. Even in rooms other than private rooms used by single individuals or workers alone, similar risks can arise if falls or unexpected accidents occur without other people around. When such falls occur, early detection and prompt treatment are necessary. However, due to privacy considerations, installing monitoring devices such as cameras in the rooms can be difficult, raising concerns about delays in detecting and treating abnormalities. For this reason, radar has traditionally been used to track the movements of living organisms and detect abnormalities based on the detected location and condition of the organism (see, for example, Patent Document 1).

[0003] Specifically, Patent Document 1 discloses that if a time variation in signal strength greater than normal occurs in the presence area (presence range bin) where the subject of measurement is present, or if the subject remains stationary for a predetermined period of time, it is determined that the subject of measurement is in an abnormal state.

[0004] Japanese Patent Application Laid-Open No. 2020-081312

[0005] In the above-mentioned Patent Document 1, the presence or absence of a subject is determined for each of a plurality of areas (range bins) divided in the distance direction of the detection area, which may increase the calculation load on the processor.

[0006] The present disclosure has been made in view of the above, and aims to realize a fall detection device and a fall detection method that reduce the calculation load.

[0007] A fall detection device according to one aspect of the present disclosure includes a sensor, a signal strength calculation unit that calculates the signal strength for each of a plurality of range bins within a detection target area based on transmission and reception signals of the sensor, a signal strength fluctuation calculation unit that calculates the signal strength fluctuation amount for each of the range bins, an area setting unit that sets a first area within the detection target area that does not include any range bins whose signal strength is equal to or greater than a signal strength threshold, and a determination unit that detects a fall of a living body within the detection target area based on a first signal strength fluctuation amount that is a statistical value of the signal strength fluctuation amount for each of the plurality of range bins within the first area.

[0008] In this configuration, a first area is set that does not include any range bins whose signal strength is equal to or greater than a signal strength threshold, and a fall of a living body within the detection target area is detected based on a first signal strength fluctuation amount, which is a statistical value of the signal strength fluctuation amount for each of multiple range bins within the first area, thereby reducing the computational load.

[0009] A fall detection method according to one aspect of the present disclosure includes calculating the signal strength for each of a plurality of range bins within a detection target area based on transmission and reception signals of a sensor, calculating a signal strength fluctuation amount for each of the range bins, setting a first area within the detection target area that does not include any range bins whose signal strength is equal to or greater than a signal strength threshold, and detecting a fall of a living body within the detection target area based on a first signal strength fluctuation amount, which is a statistical value of the signal strength fluctuation amount for each of the plurality of range bins within the first area.

[0010] In this configuration, a first area is set that does not include any range bins whose signal strength is equal to or greater than a signal strength threshold, and a fall of a living body within the detection target area is detected based on a first signal strength fluctuation amount, which is a statistical value of the signal strength fluctuation amount for each of multiple range bins within the first area, thereby reducing the computational load.

[0011] According to the present disclosure, it is possible to realize a fall detection device and a fall detection method that reduce the calculation load.

[0012] FIG. 1 is a block diagram showing a schematic configuration of a fall detection device. FIG. 2 is a schematic diagram showing a private room in which a fall detection device is installed, viewed from above. FIG. 3 is a diagram showing an example of a detection target area. FIG. 4 is a diagram showing an example of area setting when the detection target area is a private room. FIG. 5 is a diagram showing an example of area setting when the detection target area is a hospital room. FIG. 6 is a flowchart showing a specific example of fall detection processing. FIG. 7 is a block diagram showing an example of a fall detection system.

[0013] A fall detection device and a fall detection method according to embodiments will be described in detail below with reference to the drawings. However, the present disclosure is not limited to these embodiments.

[0014] 1 is a block diagram showing a schematic configuration of a fall detection device 100 according to an embodiment. The fall detection device 100 includes a radar 1, a processing unit 2, and a storage unit 3.

[0015] In the configuration of the fall detection device 100 according to the embodiment, the radar 1 is, for example, a frequency modulated continuous wave (FMCW) radar or a fast chirp modulation (FCM) radar. Since FMCW and FCM radars are well known, detailed descriptions thereof may be omitted. The present disclosure is widely applicable to sensors capable of estimating target position information using electromagnetic waves (including light) or sound waves.

[0016] The processing unit 2 is a component that can be realized by processing of an arithmetic processing device such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). The storage unit 3 is configured by a storage device such as a ROM (Read Only Memory) or a RAM (Random Access Memory).

[0017] 2 is a schematic diagram illustrating a private room in which a fall detection device is installed, viewed from above. In the example illustrated in FIG. 2, the fall detection device 100 according to the embodiment is configured and arranged to detect a person falling in the private room.

[0018] The private room in which the fall detection device 100 is installed is provided with an outward-opening door DR. In the example shown in Fig. 2, the private room is equipped with a desk 41, a chair 42, and a shelf 43. Note that the position of the door DR provided in the private room and the types and arrangement of furniture and the like installed in the private room are not limited to the example shown in Fig. 2.

[0019] 3 is a diagram showing an example of a detection target area, in which the private room shown in FIG. 2 is set as the detection target area WAR of the fall detection device 100.

[0020] 3 is the line of sight direction of the radar 1, the radar 1 may be installed in a location where transmission / reception signals can be obtained that can extract signals for each of a plurality of range bins BIN that divide the detection target area WAR in the line of sight direction (Y direction) and a direction orthogonal to the line of sight direction (X direction). While Fig. 3 shows an example in which a plurality of range bins BIN are defined based on mutually orthogonal X and Y directions, the technology disclosed herein may also be applied by defining a distance direction (R direction) and an angle direction (θ direction) with respect to the radar 1.

[0021] Specifically, the radar 1 may be installed on a wall or ceiling in a cubicle, or on structural members of furniture (desk 41, chair 42, shelf 43) installed in the cubicle. Furthermore, the radar 1 or the fall detection device 100 including the radar 1 may be placed in any location in an existing cubicle. The present disclosure is not limited by the installation or placement location of the radar 1.

[0022] The width of each range bin is, for example, 10 cm or less. The number of range bins in the detection target area WAR can be changed as appropriate depending on the width of the range bins and the size of the detection target area WAR.

[0023] The processing unit 2 includes a signal strength calculation unit 21 , a signal strength fluctuation calculation unit 22 , an area setting unit 23 , and a determination unit 24 .

[0024] The signal strength calculation unit 21 calculates a signal strength Pbin for each of a plurality of range bins BIN within the detection target area WAR based on the transmitted and received signals of the radar 1. The signal strength calculation unit 21 stores the calculated signal strength Pbin for each of the plurality of range bins BIN in the storage unit 3.

[0025] The signal strength fluctuation amount calculation unit 22 calculates the signal strength fluctuation amount PVbin for each of the plurality of range bins BIN within the detection target area WAR. The signal strength fluctuation amount calculation unit 22 stores the calculated signal strength fluctuation amount PVbin for each of the plurality of range bins BIN in the storage unit 3.

[0026] The signal strength fluctuation amount PVbin indicates the amount of time fluctuation in the signal strength Pbin for each of the multiple range bins BIN. If no living body (moving object) is present within the detection target area WAR, the signal strength fluctuation amount PVbin for each of the multiple range bins BIN within the detection target area WAR will be approximately zero.

[0027] The area setting unit 23 sets an area for determining whether or not a living body is present in the fall detection process described below, based on the signal strength calculated by the signal strength calculation unit 21. Hereinafter, the process for determining whether or not a living body is present is also referred to as a "living body presence determination process."

[0028] Specifically, the area setting unit 23 sets an area within the detection target area WAR that does not include any range bins whose signal strength Pbin is equal to or greater than the signal strength threshold as a first area PAR1, and sets an area excluding the first area PAR1 as a second area PAR2. Hereinafter, the process of setting the first area PAR1 and the second area PAR2 will also be referred to as the "area setting process." Note that the signal strength threshold used in the area setting process is set as a threshold for determining whether or not a stationary object is present in each range bin BIN within the detection target area WAR. The signal strength threshold may be a constant value that is uniform within the detection target area WAR, or may be a variable value that changes depending on the distance from the radar 1. Alternatively, the threshold determination may be performed using an adaptive algorithm such as CFAR (Constant False Alarm Rate).

[0029] The area setting process by the area setting unit 23 is performed when no living organism is present within the detection target area WAR. In other words, the area setting unit 23 performs the area setting process when the signal intensity fluctuation amount PVbin for each of the multiple range bins BIN within the detection target area WAR is approximately zero (specifically, a very small value below the noise floor in the processing unit 2).

[0030] Fig. 4 is a diagram showing an example of area setting when the detection target area is a private room. In the example shown in Fig. 4, range bins in which the signal strength Pbin exceeds the signal strength threshold are hatched. Fig. 4 shows an example in which the signal strength Pbin exceeds the signal strength threshold for a range bin in which furniture (desk 41, chair 42, shelf 43) installed in the private room and the doorknob DN of the door DR are included. As described above, in the present disclosure, an area that does not include any range bin in which the signal strength Pbin exceeds the signal strength threshold is set as a first area PAR1, and an area excluding the first area PAR1 is set as a second area PAR2.

[0031] In the fall detection device 100 according to the embodiment, the first area PAR1 is an area where it is assumed that no one will lie down under normal circumstances.

[0032] 4, if a person faints due to some kind of acute illness and falls, the person may end up lying on the line of movement between the door DR and the furniture (desk 41, chair 42, shelf 43). In this case, the space on the line of movement between the door DR and the furniture (desk 41, chair 42, shelf 43) is set as a first area PAR1, and the space including the furniture (desk 41, chair 42, shelf 43) and the doorknob DN of the door DR is set as a second area PAR2.

[0033] It should be noted that the private rooms that are the detection target area WAR of the fall detection device 100 according to the embodiment are not limited to so-called personal rooms in a typical residence, but include, for example, single-person rooms in nursing homes, individual work booths and private workspaces in businesses such as factories and warehouses, bathrooms, changing rooms, toilets, etc.

[0034] Furthermore, for example, a hospital room that can be used by multiple people at the same time can also be set as the detection target area WAR of the fall detection device 100. Fig. 5 is a diagram showing an example of area setting when the detection target area is a hospital room. In the example shown in Fig. 5, the fall detection device 100 is configured to detect a person falling within the hospital room. The example shown in Fig. 5 shows a schematic diagram of a hospital room in which the fall detection device 100 is installed, viewed from above, and illustrates an example in which the hospital room is set as the detection target area WAR of the fall detection device 100.

[0035] In the example shown in Fig. 5, range bins in which the signal strength Pbin exceeds the signal strength threshold are hatched, similar to the example shown in Fig. 4. Fig. 5 shows an example in which the signal strength Pbin exceeds the signal strength threshold in a range bin in which the signal strength Pbin exceeds the signal strength threshold, the range bin including the bed 5 and the bedside table 51.

[0036] 5, for example, it is assumed that when a patient falls from bed 5, the patient will end up lying next to the bed 5. In this case, an area including at least the space next to the bed 5 is set as a first area PAR1, and a space including the bed 5 and bedside table 51 is set as a second area PAR2.

[0037] Specifically, the first area PAR1 may include, for example, a space on a two-dimensional XY plane defined by the X and Y directions, in which at least range bins BIN whose signal strength Pbin does not exceed the signal strength threshold, that is continuous in a first direction for a first length (e.g., 60 cm) and continuous in a second direction perpendicular to the first direction for a second length (e.g., 40 cm).

[0038] 4 and 5 illustrate an example in which a space on a two-dimensional XY plane is set as an area, but a three-dimensional space including a Z direction (e.g., the height direction of a private room or hospital room) perpendicular to the XY plane may also be set as an area. As a result, the space under the desk 41 installed in the private room shown in Fig. 4 or the space on the floor under the bed 5 in the hospital room shown in Fig. 5 can be set as the first area PAR1. Alternatively, a space at a certain height or higher may be set as the second area PAR2.

[0039] In this case, the first area PAR1 may include, for example, a space in which range bins BIN whose signal strength Pbin does not exceed the signal strength threshold are continuous in a first direction on the XY plane for a first length (e.g., 60 cm), continuous in a second direction perpendicular to the first direction for a second length (e.g., 40 cm), and continuous from the floor surface for a third length (e.g., 20 cm) in the Z direction perpendicular to the XY plane.

[0040] The determination unit 24 is a component that executes various determination processes, including a process for determining the presence of a living organism in the detection target area WAR, a process for determining the presence of a living organism in the first area PAR1, and a process for determining the presence of a living organism in the second area PAR2, in the fall detection process according to the embodiment described below. A specific example of the fall detection process according to the embodiment will be described below. Fig. 6 is a flowchart showing a specific example of the fall detection process. The fall detection process shown in Fig. 6 is executed, for example, at a frame period in which the radar 1 acquires an IF signal.

[0041] The signal strength calculation unit 21 calculates the signal strength Pbin for each of the range bins BIN in the detection target area WAR based on the transmitted and received signals of the radar 1 (step S101). The signal strength calculation unit 21 stores the calculated signal strength Pbin for each of the range bins BIN in the storage unit 3.

[0042] The signal strength fluctuation amount calculation unit 22 calculates the signal strength fluctuation amount PVbin for each of the multiple range bins BIN within the detection target area WAR (step S102). Specifically, the signal strength fluctuation amount calculation unit 22 calculates the variance or standard deviation of the signal strength for multiple frames as the signal strength fluctuation amount PVbin for each of the multiple range bins BIN. Alternatively, for each of the multiple range bins BIN within the detection target area WAR, the signal strength fluctuation amount PVbin may be calculated as the difference between the signal strength Pbin(t) at time t and the signal strength Pbin(t-) at time t- before time t.

[0043] The determination unit 24 sets the maximum value of the signal strength fluctuation amount PVbin for each of the multiple range bins BIN within the detection target area WAR read from the memory unit 3 as the signal strength fluctuation amount statistical value PV(WAR) (step S103), and determines whether the signal strength fluctuation amount statistical value PV(WAR) is greater than or equal to the first threshold value PVth1 (step S104).

[0044] Here, the first threshold value PVth1 is a determination threshold value for determining whether or not a living organism is present within the detection target area WAR, and is stored in advance in the storage unit 3. Note that a minute value is set as the first threshold value PVth1 in consideration of the noise floor in the processing unit 2. Furthermore, the signal strength fluctuation amount statistical value PV(WAR) is not limited to the maximum value of the signal strength fluctuation amount PVbin for each of the multiple range bins BIN within the detection target area WAR. Specifically, the signal strength fluctuation amount statistical value PV(WAR) may be, for example, the average value of the signal strength fluctuation amount PVbin for each of the multiple range bins BIN within the detection target area WAR.

[0045] If the signal strength variation amount statistical value PV(WAR) is less than the first threshold value PVth1 (step S104; No), the determination unit 24 determines that no living organism is present within the detection target area WAR.

[0046] If the determination unit 24 determines that no living organism is present in the detection target area WAR (step S104; No), the area setting unit 23, as described above, sets the area within the detection target area WAR that does not include range bins whose signal strength Pbin is equal to or greater than the signal strength threshold as a first area PAR1, and sets the area excluding the first area PAR1 as a second area PAR2 (step S105). The area setting unit 23 associates the first area PAR1 with the range bins included in the first area PAR1, and stores them in the memory unit 3. The area setting unit 23 also associates the second area PAR2 with the range bins included in the second area PAR2, and stores them in the memory unit 3.

[0047] When the setting of the first area PAR1 and the second area PAR2 is completed (step S105), the process returns to step S101, and the processes from step S101 onward are executed in the next frame period.

[0048] If the signal strength variation amount statistical value PV(WAR) is equal to or greater than the first threshold value PVth1 (step S104; Yes), the determination unit 24 determines that a living body is present within the detection target area WAR.

[0049] If the judgment unit 24 determines that a living organism is present within the detection target area WAR (step S104; Yes), the judgment unit 24 then determines whether the first area PAR1 and the second area PAR2 have been set (step S106).

[0050] If the first area PAR1 and the second area PAR2 have not been set (step S106; No), the process returns to step S101, and the processes from step S101 onward are executed in the next frame period.

[0051] If the first area PAR1 and the second area PAR2 have been set (step S106; Yes), the judgment unit 24 then sets the maximum value of the signal strength fluctuation amount PVbin for each of the multiple range bins BIN in the second area PAR2 read from the memory unit 3 as the second signal strength fluctuation amount PV(PAR2) (step S107), and judges whether the second signal strength fluctuation amount PV(PAR2) is less than or equal to the second threshold value PVth2 (step S108).

[0052] Here, the second threshold value PVth2 is a determination threshold value for determining whether or not a living organism is present in the second area PAR2, and is stored in advance in the storage unit 3. Note that the second signal strength fluctuation amount PV(PAR2) is not limited to the maximum value of the signal strength fluctuation amount PVbin for each of the multiple range bins BIN in the second area PAR2. Specifically, the second signal strength fluctuation amount PV(PAR2) may be, for example, the average value of the signal strength fluctuation amount PVbin for each of the multiple range bins BIN in the second area PAR2.

[0053] Furthermore, when the signal intensity fluctuation statistical value PV(WAR) of the detection target area WAR and the second signal intensity fluctuation PV(PAR2) of the second area PAR2 are the same statistical value (for example, the maximum value of the signal intensity fluctuation PVbin for each of multiple range bins BIN), the first threshold value PVth1 for determining the presence of a living organism in the detection target area WAR and the second threshold value PVth2 for determining the presence of a living organism in the second area PAR2 may be the same value.

[0054] If the second signal strength variation PV(PAR2) is greater than the second threshold value PVth2 (step S108; No), the determination unit 24 determines that a living body is present in the second area PAR2.

[0055] In the example shown in Fig. 4, when a living body is present in the second area PAR2, it is assumed that, for example, a person is sitting in a chair 42 in a private room. In the example shown in Fig. 5, when a living body is present in the second area PAR2, it is assumed that, for example, a patient is lying on a bed 5 in a hospital room.

[0056] If the determination unit 24 determines that a living body is present in the second area PAR2 (step S108; No), the process returns to step S101, and the processes from step S101 onward are executed in the next frame period.

[0057] If the second signal strength fluctuation amount PV(PAR2) is equal to or less than the second threshold value PVth2 (step S108; Yes), the judgment unit 24 then sets the maximum value of the signal strength fluctuation amount PVbin for each of the multiple range bins BIN in the first area PAR1 read from the memory unit 3 as the first signal strength fluctuation amount PV(PAR1) (step S109).

[0058] The first signal strength fluctuation amount PV(PAR1) is not limited to the maximum value of the signal strength fluctuation amount PVbin for each of the multiple range bins in the first area PAR1. Specifically, the first signal strength fluctuation amount PV(PAR1) may be, for example, the average value of the signal strength fluctuation amount PVbin for each of the multiple range bins in the first area PAR1.

[0059] In the example shown in Fig. 4, for example, if a person is standing on the path of movement between the door DR and the furniture (desk 41, chair 42, shelf 43) in a private room, it is expected that the first signal strength fluctuation amount PV(PAR1) will be relatively large. Also, in the example shown in Fig. 5, for example, if a patient, doctor, or nurse is standing next to bed 5 in a hospital room, it is expected that the first signal strength fluctuation amount PV(PAR1) will be relatively large. In other words, if no abnormality has occurred in a living organism present within the detection target area WAR, it is expected that the first signal strength fluctuation amount PV(PAR1) will be relatively large.

[0060] On the other hand, in the example shown in Fig. 4, for example, if a person faints due to some acute illness in a private room and falls, lying on the path of movement between the door DR and the furniture (desk 41, chair 42, shelf 43), it is expected that the first signal intensity fluctuation amount PV(PAR1) will be relatively small. Also, in the example shown in Fig. 5, for example, if a patient in a hospital room falls off bed 5 and lies beside bed 5, it is expected that the first signal intensity fluctuation amount PV(PAR1) will be relatively small. In other words, if an abnormality occurs in a living organism present in the detection target area WAR, it is expected that the first signal intensity fluctuation amount PV(PAR1) will be relatively small.

[0061] The determination unit 24 determines whether the first signal strength fluctuation amount PV(PAR1) set in step S109 is equal to or greater than a third threshold value PVth3 (step S110).

[0062] Here, the third threshold value PVth3 is a determination threshold value for determining whether or not a living organism is present in the first area PAR1, and is stored in advance in the storage unit 3. Note that when the signal strength fluctuation amount statistical value PV(WAR) of the detection target area WAR and the first signal strength fluctuation amount PV(PAR1) of the first area PAR1 are the same statistical value (e.g., the maximum value of the signal strength fluctuation amount PVbin for each of a plurality of range bins BIN), the first threshold value PVth1 for determining that a living organism is present in the detection target area WAR and the third threshold value PVth3 for determining that a living organism is present in the first area PAR1 may be the same value.

[0063] Furthermore, if the second signal strength fluctuation amount PV(PAR2) of the second area PAR2 and the first signal strength fluctuation amount PV(PAR1) of the first area PAR1 are the same statistical value (for example, the maximum value of the signal strength fluctuation amount PVbin for each of multiple range bins BIN), the second threshold value PVth2 for determining the presence of a living organism in the second area PAR2 and the third threshold value PVth3 for determining the presence of a living organism in the first area PAR1 may be the same value.

[0064] If the first signal strength fluctuation amount PV(PAR1) is less than the third threshold value PVth3 (step S110; No), the process returns to step S101, and the processes from step S101 onward are executed in the next frame period.

[0065] If the first signal strength fluctuation amount PV(PAR1) is greater than or equal to the third threshold value PVth3 (step S110; Yes), the judgment unit 24 then judges whether the first signal strength fluctuation amount PV(PAR1) set in step S109 is less than or equal to the fourth threshold value PVth4 (step S111).

[0066] Here, the fourth threshold value PVth4 is a determination threshold value for determining whether or not an abnormality has occurred in a living organism present in the detection target area WAR, and is stored in advance in the storage unit 3. The fourth threshold value PVth4 is a value relatively larger than the third threshold value PVth3 for determining whether or not a living organism is present in the first area PAR1.

[0067] If the first signal strength fluctuation amount PV(PAR1) is greater than the fourth threshold value PVth4 (step S111; No), the process returns to step S101, and the processes from step S101 onward are executed in the next frame period.

[0068] If the first signal intensity fluctuation amount PV(PAR1) is equal to or less than the fourth threshold value PVth4 (step S111; Yes), the determination unit 24 determines that an abnormality has occurred in a living body present within the detection target area WAR. Specifically, in the example shown in FIG. 4 , for example, it is assumed that a person has fainted due to some acute illness in a private room and fallen, lying on the path of movement between the door DR and furniture (desk 41, chair 42, shelf 43). Also, in the example shown in FIG. 5 , for example, it is assumed that a patient has fallen off bed 5 in a hospital room and is lying next to bed 5.

[0069] In the present disclosure, when the determination unit 24 determines that an abnormality has occurred in a living organism present within the detection target area WAR, it detects that a person has fallen within the detection target area WAR. When the determination unit 24 detects that a person has fallen within the detection target area WAR, it issues an alarm to an external device indicating that a person has fallen within the detection target area WAR (step S112), and ends the fall detection process shown in FIG.

[0070] 7 is a block diagram showing an example of a fall detection system. The fall detection system 300 according to the embodiment includes the above-described fall detection device 100 and an external device 200.

[0071] The fall detection device 100 and the external device 200 are configured to be able to communicate with each other. The communication means between the fall detection device 100 and the external device 200 may be, for example, a wireless communication means defined by a communication standard such as 4G (fourth generation mobile communication system), 5G (fifth generation mobile communication system), LTE (Long Term Evolution)-FDD (Frequency Division Duplex), LTE-TDD (Time Division Duplex), LTE-Advanced, or LTE-Advanced Pro, or the fall detection device 100 and the external device 200 may be connected by wire.

[0072] The external device 200 may, for example, display an abnormality on a monitor (not shown) provided in a room other than the private room (see FIG. 4) or hospital room (see FIG. 5) in which the fall detection device 100 is installed, or may sound an alarm from a speaker (not shown). Furthermore, the external device 200 may, for example, be a mobile terminal device such as a smartphone of the person being monitored by the fall detection device 100.

[0073] In the fall detection process described above, the area setting process is executed in a state where no living body is present within the detection target area WAR.

[0074] Specifically, if the signal strength fluctuation statistical value PV(WAR) of the detection target area WAR is less than the first threshold value PVth1 (step S104; No), the fall detection device 100 determines that no moving living body is present within the detection target area WAR and executes area setting processing (step S105).If the signal strength fluctuation statistical value PV(WAR) of the detection target area WAR is equal to or greater than the first threshold value PVth1 (step S104; Yes), the fall detection device 100 executes living body presence determination processing for each of the first area PAR1 and the second area PAR2.

[0075] This means that, for example, even if the installation location of radar 1 or the fall detection device 100 including radar 1 is changed, it is possible to perform the living body presence determination process for each of the appropriately set first area PAR1 and second area PAR2.

[0076] In addition, in the fall detection process described above, an area where it is expected that a person will not normally lie down is set as the first area PAR1, and a fall of a living body within the detection target area WAR is determined based on statistical values ​​such as the maximum and average values ​​of the signal intensity fluctuation amount PVbin for each of multiple range bins BIN within the first area PAR1.

[0077] Specifically, in the process of determining the presence of a living organism in the first area PAR1, the fall detection device 100 sets a statistical value such as the maximum or average value of the signal strength fluctuation amount PVbin for each of the multiple range bins BIN in the first area PAR1 as the first signal strength fluctuation amount PV(PAR1) (step S109).If the first signal strength fluctuation amount PV(PAR1) of the first area PAR1 is equal to or greater than a third threshold value PVth3 for determining whether a living organism is present in the first area PAR1 (step S110; Yes), and if the first signal strength fluctuation amount PV(PAR1) of the first area PAR1 is equal to or less than a fourth threshold value PVth4 for determining whether an abnormality has occurred in the living organism present in the detection target area WAR (step S111; Yes), the fall detection device 100 determines that an abnormality has occurred in the living organism present in the detection target area WAR, and detects that a human has fallen in the detection target area WAR.

[0078] This can reduce the calculation load compared to, for example, a mode in which the signal intensity fluctuation amount PVbin for each of a plurality of range bins BIN in the detection target area WAR is threshold-determined to detect an abnormality in a living body.

[0079] For example, in the area setting process, a machine learning model may be generated from the signal intensity distribution of the detection target area WAR, with classification data corresponding to the first area PAR1 and the second area PAR2 used as teacher labels for each range bin, and the acquired signal intensity distribution of the detection target area WAR may be used as input to set the first area PAR1 and the second area PAR2 for each range bin.

[0080] Furthermore, for example, in the process of determining the presence of a living organism, a machine learning model may be generated from the signal intensity fluctuation distribution of the detection target area WAR, using classification data corresponding to the presence of a living organism and the occurrence of an abnormality as teacher labels for each range bin, and the acquired signal intensity fluctuation distribution of the detection target area WAR may be used as input to determine the presence of a living organism and the occurrence of an abnormality.

[0081] The above-described embodiments are intended to facilitate understanding of the present disclosure and are not intended to limit the present disclosure. The present disclosure may be modified or improved without departing from the spirit thereof, and equivalents thereof are also included in the present disclosure.

[0082] The present disclosure can have the following configurations as described above or instead of the above.

[0083] (1) A fall detection device according to one aspect of the present disclosure includes a sensor, a signal strength calculation unit that calculates the signal strength for each of a plurality of range bins within a detection target area based on transmission and reception signals of the sensor, a signal strength fluctuation calculation unit that calculates the signal strength fluctuation amount for each of the range bins, an area setting unit that sets a first area within the detection target area that does not include any range bins whose signal strength is equal to or greater than a signal strength threshold, and a determination unit that detects a fall of a living body within the detection target area based on a first signal strength fluctuation amount, which is a statistical value of the signal strength fluctuation amount for each of the plurality of range bins within the first area.

[0084] In this configuration, a first area is set that does not include any range bins whose signal strength is equal to or greater than a signal strength threshold, and a fall of a living body within the detection target area is detected based on a first signal strength fluctuation amount, which is a statistical value of the signal strength fluctuation amount for each of multiple range bins within the first area, thereby reducing the computational load.

[0085] (2) In the fall detection device of (1) above, the sensor is preferably an FMCW (Frequency Modulated Continuous Wave) radar or an FCM (Fast Chirp Modulation) radar.

[0086] (3) In the fall detection device of (1) or (2) above, the area setting unit sets the first area when a signal strength fluctuation amount statistical value, which is a statistical value of the signal strength fluctuation amount for each of a plurality of range bins within the detection target area, is less than a first threshold value.

[0087] This reduces the computational load when setting the first area.

[0088] (4) In the fall detection device of (3) above, the signal strength fluctuation amount statistical value is a maximum value or an average value of the signal strength fluctuation amount for each of a plurality of bins within the detection target area.

[0089] (5) In the fall detection device of (3) above, the first signal strength fluctuation amount is a maximum value or an average value of the signal strength fluctuation amount for each of a plurality of bins in the first area.

[0090] (6) In the fall detection device of (3) above, the area setting unit sets a second area excluding the first area.

[0091] (7) In the fall detection device of (6) above, the determination unit detects a fall of a living body within the detection target area based on the first signal strength fluctuation amount when the signal strength fluctuation amount statistical value is equal to or greater than a first threshold value.

[0092] This reduces the computational load when detecting a fall of a living body within a detection target area.

[0093] (8) In the fall detection device of (7) above, the determination unit determines that a fall of a living body has occurred within the detection target area when the second signal intensity fluctuation amount, which is a statistical value of the signal intensity fluctuation amount for each of multiple range bins within the second area, is less than or equal to a second threshold, and the first signal intensity fluctuation amount is greater than or equal to a third threshold, and the first signal intensity fluctuation amount is less than or equal to a fourth threshold that is greater than the third threshold.

[0094] This reduces the computational load when detecting a fall of a living body within a detection target area.

[0095] (9) In the fall detection device of (8) above, the second signal strength fluctuation amount is a maximum value or an average value of the signal strength fluctuation amount for each of a plurality of bins in the second area.

[0096] (10) A fall detection method according to one aspect of the present disclosure includes a fall detection device according to any one of (1) to (9) above and an external device configured to be able to communicate with the fall detection device, wherein the fall detection device issues an alarm to the external device indicating that it has determined that a fall has occurred within the detection target area.

[0097] In this configuration, it is possible to notify an external device that a fall of a living body has been detected within the detection target area.

[0098] The present disclosure makes it possible to realize a fall detection device and a fall detection method that reduce the computational load.

[0099] REFERENCE SIGNS LIST 1 radar 2 processing unit 3 storage unit 5 bed 21 signal strength calculation unit 22 signal strength variation amount calculation unit 23 area setting unit 24 determination unit 41 desk 42 chair 43 shelf 51 bedside table 100 fall detection device 200 external device 300 fall detection system

Claims

1. A fall detection device comprising: a sensor; a signal strength calculation unit that calculates the signal strength for each of a plurality of range bins within a detection target area based on transmission and reception signals of the sensor; a signal strength fluctuation amount calculation unit that calculates the signal strength fluctuation amount for each of the range bins; an area setting unit that sets a first area within the detection target area that does not include any range bins whose signal strength is equal to or greater than a signal strength threshold; and a determination unit that detects a fall of a living body within the detection target area based on a first signal strength fluctuation amount, which is a statistical value of the signal strength fluctuation amount for each of the plurality of range bins within the first area.

2. A fall detection device according to claim 1, wherein the sensor is an FMCW (Frequency Modulated Continuous Wave) radar or an FCM (Fast Chirp Modulation) radar.

3. A fall detection device according to claim 1 or 2, wherein the area setting unit sets the first area when a signal strength fluctuation amount statistical value, which is a statistical value of the signal strength fluctuation amount for each of a plurality of range bins within the detection target area, is less than a first threshold value.

4. A fall detection device according to claim 3, wherein the signal strength fluctuation amount statistical value is the maximum value or average value of the signal strength fluctuation amount for each of a plurality of range bins within the detection target area.

5. A fall detection device according to claim 3, wherein the first signal strength fluctuation amount is the maximum value or average value of the signal strength fluctuation amount for each of a plurality of range bins in the first area.

6. A fall detection device according to claim 3, wherein the area setting unit sets a second area excluding the first area.

7. A fall detection device according to claim 6, wherein the determination unit detects the occurrence of a fall of a living body within the detection target area based on the first signal strength fluctuation amount when the signal strength fluctuation amount statistical value is equal to or greater than a first threshold value.

8. A fall detection device as described in claim 7, wherein the determination unit determines that a fall of a living body has occurred within the detection target area when a second signal intensity fluctuation amount, which is a statistical value of the signal intensity fluctuation amount for each of a plurality of range bins within the second area, is equal to or less than a second threshold, and the first signal intensity fluctuation amount is equal to or greater than a third threshold, and the first signal intensity fluctuation amount is equal to or less than a fourth threshold that is greater than the third threshold.

9. A fall detection device according to claim 8, wherein the second signal strength fluctuation amount is a maximum value or an average value of the signal strength fluctuation amount for each of a plurality of range bins in the second area.

10. A fall detection device according to any one of claims 1 to 9, wherein the determination unit issues an alarm to an external device indicating that it has determined that a fall has occurred within the detection target area.

11. A fall detection method comprising: calculating signal strength for each of a plurality of range bins within a detection target area based on transmission and reception signals of a sensor; calculating a signal strength fluctuation amount for each of the range bins; setting a first area within the detection target area that does not include any range bins whose signal strength is equal to or greater than a signal strength threshold; and detecting a fall of a living body within the detection target area based on a first signal strength fluctuation amount, which is a statistical value of the signal strength fluctuation amount for each of the plurality of range bins within the first area.

12. A fall detection method according to claim 11, wherein the sensor is an FMCW (Frequency Modulated Continuous Wave) radar or an FCM (Fast Chirp Modulation) radar.

13. A fall detection method according to claim 11 or 12, wherein the first area is set when a signal intensity fluctuation amount statistical value, which is a statistical value of the signal intensity fluctuation amount for each of a plurality of range bins within the detection target area, is less than a first threshold value.

14. A fall detection method according to claim 13, wherein the signal strength fluctuation amount statistical value is the maximum value or average value of the signal strength fluctuation amount for each of a plurality of range bins within the detection target area.

15. A fall detection method according to claim 13, wherein the first signal strength fluctuation amount is the maximum value or average value of the signal strength fluctuation amount for each of a plurality of range bins in the first area.

16. A fall detection method according to claim 13, further comprising setting a second area excluding the first area.

17. A fall detection method as claimed in claim 16, wherein, when the signal strength fluctuation amount statistical value is equal to or greater than a first threshold value, a fall of a living body within the detection target area is detected based on the first signal strength fluctuation amount.

18. A fall detection method as described in claim 17, wherein a fall of a living body is determined to have occurred within the detection target area when a second signal intensity fluctuation amount, which is a statistical value of the signal intensity fluctuation amount for each of a plurality of range bins within the second area, is equal to or less than a second threshold, and the first signal intensity fluctuation amount is equal to or greater than a third threshold, and the first signal intensity fluctuation amount is equal to or less than a fourth threshold that is greater than the third threshold.

19. A fall detection method according to claim 18, wherein the second signal strength fluctuation amount is the maximum value or average value of the signal strength fluctuation amount for each of a plurality of range bins in the second area.

20. A fall detection method according to any one of claims 11 to 19, wherein an alarm indicating that a fall has been determined to have occurred within the detection target area is issued to an external device.

Images