Living body detection device

Abstract

According to the present invention, a transmission / reception unit receives, by means of a plurality of reception antennas, reflected waves of radio waves transmitted from at least one transmission antenna into an observation range, and generates an intermediate frequency signal based on the transmission signal and the reception signal. A processing unit counts the positions of living bodies and the number of the living bodies within the observation range on the basis of the intermediate frequency signal. On the basis of the intermediate frequency signal, the processing unit calculates a complex signal reflecting the presence or absence of an object and the speed of the object for each position within the observation range. Furthermore, on the basis of the complex signal, a living body determination index is calculated for each position within the observation range the living body determination index varying in magnitude depending on whether the detected object is a living body and on the number of the living bodies. The living body determination index is used to identify the positions of the living bodies within the observation range and to count the number of the living bodies.

Description

Biological detection device

[0001] This invention relates to a biological detection device.

[0002] A device for detecting living organisms using radar is known (Patent Document 1). The device described in Patent Document 1 receives multiple reflected signals reflected by at least one person within the observation range. Based on these multiple reflected signals, multiple vital signals relating to at least one person within the observation range are acquired, and the similarity between the multiple vital signals is calculated. Based on the similarity, it is determined whether the multiple vital signals relate to one person or to multiple people. This suppresses false detection of the number of people.

[0003] Japanese Patent Publication No. 2024-128762

[0004] When multiple individuals are in close proximity, below the angular resolution of the radar, the similarity of their vital signals increases due to the influence of multipath waves passing through them. As a result, even though multiple people are present, they may be mistakenly identified as a single individual. The objective of the present invention is to provide a biological detection device that can improve the accuracy of counting the number of living organisms within the observation range.

[0005] According to one aspect of the present invention, a biological detection device is provided comprising: a transmitting and receiving unit that receives reflected waves of radio waves transmitted from at least one transmitting antenna to an observation range with a plurality of receiving antennas and generates a transmitting signal and an intermediate frequency signal based on the receiving signal; and a processing unit that identifies the location of a living organism within the observation range and counts the number of living organisms based on the intermediate frequency signal, wherein the processing unit calculates a complex signal that reflects the presence or absence of an object and the velocity of the object for each location within the observation range based on the intermediate frequency signal, calculates a biological determination index for each location within the observation range that differs in size depending on whether the detected object is a living organism or not, and whose size depends on the number of living organisms, based on the complex signal, and identifies the location of a living organism within the observation range and counts the number of living organisms based on the biological determination index.

[0006] The number of living organisms can be counted from the size of a biological identification index whose size depends on the number of living organisms. Therefore, the accuracy of counting multiple living organisms in close proximity can be improved compared to the spatial resolution of object detection.

[0007] Figure 1 is a block diagram of the biological detection device according to the first embodiment. Figure 2 is a flowchart showing the processing procedure executed by the transmitting / receiving unit 20 and the processing unit 30. Figure 3 is a schematic diagram showing the arrangement of various information (data) obtained in the processing procedure of step S3 (Figure 2). Figure 4 is a flowchart showing the detailed processing procedure of step S5 (Figure 2). Figure 5 is a graph showing an example of the relationship between the number of living organisms and the size of the biological determination index I. Figure 6A is a schematic plan view showing the arrangement of objects within the observation range 50, Figure 6B is a diagram showing the power distribution obtained by squaring the absolute value of the complex signal obtained in step S3 (Figure 2) using shades of gray, and Figure 6C is a diagram showing the distribution of the biological determination index I obtained in step S4 (Figure 2) using shades of gray. Figure 7 is a flowchart showing the processing procedure executed by the transmitting / receiving unit 20 and the processing unit 30 of the biological detection device according to the second embodiment. Figure 8 shows the power distribution obtained by squaring the absolute value of the complex signal obtained in step S3 (Figure 7) using shades of gray, and also shows the state in which the observation range 50 is divided into four sections D1, D2, D3, and D4. Figure 9 is a schematic diagram showing the arrangement of various information (data) that appears in the procedure for obtaining the amplitude component m(x, y, t) and phase component p(x, y, t) from the three-dimensional data X(r, d, n) in the third embodiment. Figure 10 is a flowchart showing the processing procedure executed by the transmitting / receiving unit 20 and the processing unit 30 of the biodetector according to the third embodiment. Figure 11 is a flowchart showing the detailed procedure of step SB5 (Figure 10). Figure 12 is a schematic diagram showing an example of a point cloud 40 distributed in the observation range 50, i.e., the floor surface (xy plane). Figures 13A to 13E are graphs showing the time evolution of various features of the cluster. Figure 14A is a schematic diagram showing an example of the arrangement of clusters 45A and 45B when two clusters 45A and 45B are defined within the observation range 50 in step SB52 (Figure 11). Figure 14B is a schematic diagram illustrating an example of a method for determining the location of a living organism. Figure 14C is a schematic diagram illustrating another example of a method for determining the location of a living organism. Figure 15 is a schematic diagram illustrating yet another example of a method for determining the location of a living organism.Figure 16 is a schematic diagram showing the arrangement of various information (data) that appears in the procedure for obtaining the amplitude component m(x, y, z) and phase component p(x, y, z), which are three-dimensional data, from the three-dimensional data X(r, d, n) in the fourth embodiment. Figure 17 is a schematic diagram showing the arrangement of various information (data) that appears in the procedure for obtaining the complex signal P(r, θ, φ) from the three-dimensional data X(r, d, n). Figure 18 is a schematic diagram showing the arrangement of various information (data) that appears in the procedure for obtaining the complex signal P'(r, θ, φ) originating from the living organism from the biological judgment index I'(r, d, n).

[0008] [First Embodiment] A biological detection device according to the first embodiment will be described with reference to the drawings from Figure 1 to Figure 6C. Figure 1 is a block diagram of the biological detection device according to the first embodiment. The biological detection device according to the first embodiment includes a transmitting / receiving unit 20 and a processing unit 30.

[0009] The transmitting / receiving unit 20 includes a signal generating unit 22, a plurality of AD converters 23, a plurality of mixers 24, at least one transmitting antenna Tx, and a plurality of receiving antennas Rx.

[0010] The signal generation unit 22 periodically supplies a transmission signal called a chirp to the transmitting antenna Tx, which linearly increases in frequency over time. The transmitting antenna Tx transmits radio waves within the observation range 50. Each of the multiple receiving antennas Rx receives reflected waves from objects 51 within the observation range 50.

[0011] Mixers 24, each prepared for multiple receiving antennas Rx, mix the transmitted signal and the received signal to generate an intermediate frequency signal (IF signal). AD converters 23 perform AD conversion on each of the IF signals. The AD-converted IF signals are input to the processing unit 30.

[0012] The processing unit 30 estimates the position and velocity of objects within the observation range 50 based on the IF signal, determines whether the detected object 51 is a living organism or not, and counts the number of living organisms within the observation range 50.

[0013] Next, the functions of the processing unit 30 will be explained with reference to Figure 2. Figure 2 is a flowchart showing the processing procedure executed by the processing unit 30.

[0014] First, the transmitting / receiving unit 20 transmits and receives signals (step S1). The transmitted signal is a frequency continuous modulation wave in which chirps, in which the frequency changes linearly with respect to time, repeatedly appear. The mixer 24 and the AD converter 23 generate an IF signal for each channel, which is composed of multiple pairs of transmitting antennas Tx and receiving antennas Rx (step S2).

[0015] Next, the processing unit 30 performs distance FFT, velocity FFT, and angle estimation processing on the IF signal to calculate the distribution of complex signals within the observation range 50 and the power obtained by squaring the absolute value of these complex signals (step S3). This processing is the same as that of a known FMCW radar system. For example, distance FFT is performed for each chirp, and velocity FFT is performed on the result of the distance FFT processing for each chirp frame consisting of multiple chirps. Angle estimation processing is performed on the result of the velocity FFT processing for multiple channels. The calculated complex signals are represented by complex numbers that reflect the presence or absence of objects and the velocity of objects at each position within the observation range 50, and the power is represented by the square of the absolute value of the complex signal.

[0016] The process of step S3 will be explained with reference to Figure 3. Figure 3 is a schematic diagram showing the arrangement of various information (data) obtained in the processing procedure of step S3. By performing distance FFT processing and velocity FFT processing, three-dimensional data X(r,d,n) defined in the three-dimensional space of distance r, velocity d, and channel n is calculated.

[0017] Next, the processing unit 30 averages the three-dimensional data X in the velocity direction. This gives the two-dimensional data X'(r, n) defined in a two-dimensional space of distance r and channel n. By performing angle estimation processing based on the two-dimensional data X'(r, n), a complex signal P(r, θ) defined in a two-dimensional space of distance r and angle θ is calculated. The complex signal P is a complex number and includes an amplitude component m and a phase component p. The complex signal P reflects the presence or absence of an object and the velocity of the object. The position of an object can be detected from the complex signal P or the power distribution. The complex signal P(r, θ) can be considered as a signal that reflects the amplitude and phase of the reflected wave from position (r, θ).

[0018] The processing unit 30 calculates the amplitude component m(r, θ) and phase component p(r, θ) of the complex signal P(r, θ). By calculating the amplitude component m(r, θ) and phase component p(r, θ) for each of the multiple chirp frames and saving the calculation results, the amplitude component m(r, θ, t) and phase component p(r, θ, t) of the complex signal P can be obtained for each distance r, angle θ, and discretized time t.

[0019] In step S3 (Figure 2), once the complex signal distribution, for example, the amplitude component m(r,θ,t) and the phase component p(r,θ,t), is determined, the processing unit 30 calculates a biological determination index I(r,θ) for each position (r,θ) (step S4).

[0020] Next, we will explain the biological identification index I(r,θ). The surface of a living organism is periodically displaced by respiration, heartbeat, etc. When a living organism is present at a specific location (r,θ), the amplitude component m(r,θ,t) and phase component p(r,θ,t) at that location largely reflect characteristics caused by the periodic displacement of the body surface. For example, the time change of the amplitude component m(r,θ,t) and the time change of the phase component p(r,θ,t) begin to show the same trend, and the correlation between the two becomes strong. The biological identification index I(r,θ) is defined such that when the characteristics caused by the periodic displacement of the body surface are largely reflected in the amplitude component m(r,θ,t) and phase component p(r,θ,t), the biological identification index I(r,θ) becomes large.

[0021] When a stationary object other than a living organism is present at a specific position (r, θ), the amplitude component m(r, θ, t) and phase component p(r, θ, t) do not exhibit features caused by the periodic displacement of the body surface. Random noise components are superimposed on the time evolution of the amplitude component m(r, θ, t) and phase component p(r, θ, t), thus weakening the correlation between the two.

[0022] Furthermore, the amplitude of the time variation of the amplitude component m caused by displacement of the body surface is greater than the amplitude of the time variation of the noise component. For this reason, the variance of the amplitude component m(r, θ, t) when a living organism is present at a specific position (r, θ) is greater than the variance of the amplitude component m(r, θ, t) when a stationary object is present.

[0023] Using the correlation between the temporal changes of the amplitude component m(r, θ, t) and the phase component p(r, θ, t), and the properties of the variance of the amplitude component m(r, θ, t) as described above, it is possible to determine whether the detected object is a living body or not.

[0024] The strength C of the correlation between the temporal changes of the amplitude component m(r, θ, t) and the phase component p(r, θ, t) ｍｐ is defined by the following equation. is the time width for calculating the strength of the correlation, m_overbar is the average of the amplitude component m(r, θ, t) from time 0 to T, and p_overbar is the average of the phase component p(r, θ, t) from time 0 to T.

[0025] The variance σ of the amplitude component m(r, θ, t) ｍ ２ can be calculated, for example, by the following equation.

[0026] The living body determination index I(r, θ, t) is defined, for example, by the following equation.

[0027] When the living body determination index I(r, θ, t) is calculated in step S4 (Figure 2), the processing unit 30 identifies the position of the living body using the living body determination index I(r, θ, t) and counts the number of living bodies. For example, the processing unit 30 identifies the position of the living body and counts the number of living bodies based on the spatial distribution of the living body determination index I(r, θ, t) and the sum of the living body determination index I(r, θ, t) within the observation range 50 (step S5).

[0028] Next, referring to Figure 4, the detailed processing procedure of step S5 (Figure 2) will be described. Figure 4 is a flowchart showing the detailed processing procedure of step S5.

[0029] The processing unit 30 identifies the location of living organisms from the distribution of the biological detection index I(r,θ,t) and counts the number of living organisms (step S51). For example, it detects the peak position of the distribution of the biological detection index I(r,θ,t). If the peak value exceeds a predetermined judgment threshold, it can be estimated that a living organism is present at that peak position. This judgment threshold may be set to a value sufficiently larger than the average value of the biological detection index I in areas where no living organisms exist, for example, 100 times that value, after obtaining the average value in advance. Alternatively, the judgment threshold may be set by applying the known CA-CFAR (Cell-Averaging Constant False Alarm Rate) algorithm. The number of living organisms can be counted by counting the number of peaks.

[0030] Next, the processing unit 30 calculates the sum of the biological determination index I in the observation range 50 (step S52). Then, based on the calculated sum of the biological determination index I, the number of organisms counted in step S51 is corrected (step S53). In the organism counting process in step S51, if multiple organisms are located close together compared to the spatial resolution of the peak position detection of the distribution of biological determination index I, the number of organisms is counted as 1. In step S53, the number of organisms located at a single peak position is counted more accurately.

[0031] Next, with reference to Figure 5, we will explain the method for correcting the number of living organisms. As shown in equation (3), the correlation C ｍｐ The variance becomes stronger, ｍ ２ As the value of increases, the biological identification index I also increases. The biological identification index I is larger when there are multiple organisms at the peak position than when there is one. By conducting various evaluation experiments in advance, the relationship between the number of organisms and the magnitude of the biological identification index I can be determined.

[0032] Figure 5 is a graph showing an example of the relationship between the number of living organisms and the magnitude of the biological determination index I. The horizontal axis represents the sum of the biological determination index I, and the vertical axis represents the number of living organisms. Various evaluation experiments have shown that the sum of the biological determination index I is approximately proportional to the number of living organisms. For example, when the number of living organisms is 1, the sum of the biological determination index I is approximately 2, and when the number of living organisms is 2, the sum of the biological determination index I is approximately 4. If the relationship shown in Figure 5 is known in advance, the number of living organisms can be determined from the sum of the biological determination index I. In step S53, the number of living organisms counted in step S51 is corrected to the number of living organisms determined based on the relationship shown in Figure 5.

[0033] Next, with reference to Figures 6A, 6B, and 6C, the results of an evaluation experiment to detect living organisms within the observation range 50 will be described. Figure 6A is a schematic plan view showing the arrangement of objects within the observation range 50. A biological detection device 55 is positioned approximately in the center of one of the shorter sides of the rectangular observation range 50. Two people 53 are located in close proximity in front of the biological detection device 55, and a stationary object 52 is located at approximately the same distance as the people 53. A wall 54 is located further away from the people 53 and the stationary object 52.

[0034] Figure 6B is a diagram showing the power distribution obtained from the square of the absolute value of the complex signal obtained in step S3 (Figure 2), using shades of gray. Regions with relatively high power are shown in darker shades. In regions 53A, 52A, and 54A, which correspond to the person 53, the stationary object 52, and the wall 54, the power is relatively high. At this stage, since it has not been determined whether the objects are living organisms or not, the person 53, the stationary object 52, and the wall 54 are detected.

[0035] Figure 6C shows the distribution of the biological detection index I obtained in step S4 (Figure 2) using shades of gray. The biological detection index I is smaller in the regions 52A and 54A (Figure 6B) corresponding to the stationary object 52 and the wall 54, while the biological detection index I is relatively larger in the region 53A corresponding to the person 53. This suppresses the complex signal P (Figure 3) and its power originating from the stationary object 52, etc., making it possible to detect a living organism.

[0036] However, since the peak position of the distribution of the biological determination index I shown in FIG. 6C is at one location, in step S51 (FIG. 4), the number of living bodies is counted as 1. Thereafter, by executing steps S52 and S53, the number of living bodies is corrected to 2. In this way, the number of living bodies can be accurately counted.

[0037] Next, the excellent effects of the first embodiment will be described. In the first embodiment, by correcting the number of living bodies in step S53 (FIG. 4), even when a plurality of living bodies are present in proximity, the counting accuracy of the living bodies can be improved.

[0038] Next, a modified example of the first embodiment will be described. In the first embodiment, as shown in formula (3), as the biological determination index I for determining whether an object is a living body, the square of the absolute value (power) of the complex signal P, the strength c of the correlation ｍｐ , and the variance σ ｍ ２ are used for calculation. However, as the biological determination index I, instead of using the square of the absolute value (power) of the complex signal P, the strength c of the correlation ｍｐ and the variance σ ｍ ２ may be used.

[0039] Furthermore, the reason for using the variance σ ｍ ２ in the biological determination index I is that attention is paid to the fact that the variance of the amplitude caused by the displacement of the body surface of the living body is larger than the variance of the amplitude of the random noise. In order to determine whether an object is a living body, the variance σ ｍ ２ is not essential. For example, only the strength c of the correlation ｍｐ may be adopted as the biological determination index I.

[0040] In addition, as the biological determination index I, for each position within the observation range 50, an index whose size varies depending on whether the detected object is a living body and whose size depends on the number of living bodies may be adopted.

[0041] The above embodiment shows an example of detecting a living organism using an FMCW radar, but other types of radar may be used. For example, Doppler radar, pulse radar, OFDM (Orthogonal Frequency Division Multiplexing) radar, etc., may be used. In this case, among the various pieces of information obtained from the received signal, an index that reflects the periodic displacement of the body surface caused by the respiration and heartbeat of the living organism may be used as the biological identification index.

[0042] [Second Embodiment] Next, a biological detection device according to the second embodiment will be described with reference to Figures 7 and 8. Hereinafter, the configuration common to the biological detection device according to the first embodiment, which was described with reference to Figures 1 to 6C, will not be explained.

[0043] Figure 7 is a flowchart showing the processing procedures performed by the transmitting / receiving unit 20 and the processing unit 30 of the biological detection device according to the second embodiment. The procedures from step S1 to step S4 are the same as the procedures from step S1 to step S4 in the first embodiment (Figure 2). When a biological determination index I is calculated at each position within the observation range 50, the processing unit 30 uses the biological determination index I to identify the location of the biological organism. For example, the processing unit 30 identifies the location of the biological organism based on the distribution of the biological determination index I (step SA5). Subsequently, the processing unit 30 divides the observation range 50 into multiple sections (step SA6).

[0044] Figure 8 shows the distribution of the biological detection index I obtained in step SA5 (Figure 7) using shades of gray, and also illustrates the state in which the observation range 50 is divided into four sections D1, D2, D3, and D4. The area 53A corresponding to the person 53 (Figure 6A) is located within section D4. When dividing into multiple sections, the sections are determined so that the boundary lines of the sections do not pass through the location where a biological organism is detected, for example, the dark gray area 53A.

[0045] The processing unit 30 divides the observation range 50 into multiple sections, and then counts the number of living organisms in each section using the biological detection index I. For example, the processing unit 30 counts the number of living organisms based on the sum of the biological detection index I in each section (step SA7). The process in step SA7 is the same as the process in step S53 (Figure 4) performed by the biological detection device according to the first embodiment. Step SA7 is repeated until the counting process is completed for all sections (step SA8).

[0046] In the example shown in Figure 8, the count of living organisms is zero in sections D1, D2, and D3, and the count of living organisms is 2 in section D4.

[0047] Next, the excellent effects of the second embodiment will be described. In the second embodiment, as in the first embodiment, the accuracy of counting living organisms can be improved even when multiple living organisms are present in close proximity.

[0048] If the IF signal generated in step S2 contains disturbances, noise is superimposed on the biological detection index I. When the range over which the sum of the biological detection index I is calculated is wide, more noise is added to the sum. As a result, the accuracy of the count result for the number of living organisms decreases. As in the second embodiment, when the range over which the sum of the biological detection index I is calculated is narrowed, the sum becomes less susceptible to noise, and thus the decrease in the accuracy of the count result is suppressed.

[0049] Furthermore, in the second embodiment, the number of living organisms in each section can be counted. For example, if living organisms are detected in two locations within the observation range 50, in the first embodiment, the total number of living organisms present in the two locations can be counted, but the number of living organisms in each of the two locations cannot be counted individually. In the second embodiment, if the two locations where living organisms are detected are divided into different sections, the number of living organisms in each of the two locations can be counted individually.

[0050] [Third Embodiment] Next, a biological detection device according to the third embodiment will be described with reference to Figures 9 to 15. Hereinafter, the configuration common to the biological detection device according to the first embodiment, which was described with reference to Figures 1 to 6C, will be omitted from the explanation.

[0051] In the first embodiment, as shown in Figure 6A, radio waves are emitted almost horizontally from a biodetector 55 mounted on the wall of the room. In the third embodiment, however, radio waves are emitted toward the floor from a biodetector 55 mounted on the ceiling of the room, for example. In the third embodiment, the observation range is the floor, and the position and number of living organisms present on the floor are detected. The biodetection method according to the third embodiment can also be applied when radio waves are emitted horizontally, as illustrated in the first embodiment.

[0052] Figure 9 is a schematic diagram showing the arrangement of various pieces of information (data) that appear in the procedure for obtaining the amplitude component m(x,y,t) and phase component p(x,y,t) of a complex signal P(x,y) from three-dimensional data X(r,d,n). The three-dimensional data X(r,d,n) is the same as the three-dimensional data X(r,d,n) shown in Figure 3 of the first embodiment. In the first embodiment, as shown in Figure 3, the amplitude component m(r,θ,t) and phase component p(r,θ,t) are defined in polar coordinates, but in the third embodiment, the amplitude component m(x,y,t) and phase component p(x,y,t) are defined in the xy Cartesian coordinate system.

[0053] Furthermore, in order to acquire information within the floor surface of the observation range, the transmitting antenna Tx and the receiving antenna Rx are arranged in two dimensions. For example, at least one of the vectors from the transmitting antenna Tx to the receiving antenna Rx that constitute each of the multiple channels has an x-direction component parallel to the floor surface, and at least one other has a y-direction component parallel to the floor surface and perpendicular to the x-direction.

[0054] Similar to the first embodiment, the three-dimensional data X(r,d,n) is averaged in the velocity direction (d direction) to obtain the two-dimensional data X'(r,n). An angle estimation process is performed based on the two-dimensional data X'(r,n) to calculate a complex signal P(r,θ,φ) defined in three-dimensional space, represented by distance r, angle with respect to the z axis (polar angle) φ, and azimuth angle θ in the xy plane. The complex signal P(r,θ,φ) is accumulated in the φ direction to calculate a complex signal P(r,θ) defined in two-dimensional space. The complex signal P(r,θ) is converted into a complex signal P(x,y) expressed in xy Cartesian coordinates. Here, the x and y directions are parallel to the floor.

[0055] The amplitude component m(x,y) and phase component p(x,y) of the complex signal P(x,y) are calculated. The amplitude component m(x,y) and phase component p(x,y) are calculated for each of the multiple chirp frames. Through these processes, the amplitude component m(x,y,t) and phase component p(x,y,t) are obtained for each discretized time t.

[0056] Figure 10 is a flowchart showing the processing procedure performed by the transmitting / receiving unit 20 (Figure 1) and processing unit 30 (Figure 1) of the biological detection device according to the third embodiment. The procedure from step S1 to step S4 is the same as the procedure from step S1 to step S4 of the first embodiment (Figure 3). The complex signal and power distribution calculated in step S3 are defined in the xy plane parallel to the floor surface, as shown in Figure 9. In the first embodiment, the biological determination index I calculated in step S4 is defined in polar coordinates (r, θ) and time t, as shown in equation (3), but in the third embodiment, the biological determination index I is defined in xy coordinates (x, y) and time t.

[0057] In the third embodiment, the location of a living organism is identified and the number of living organisms is counted using the power of a biological identification index I(x,y,t) and a complex signal P(x,y).

[0058] Referring to Figure 11, the detailed procedure of step SB5 (Figure 10) will be described. Figure 11 is a flowchart showing the detailed procedure of step SB5. First, the processing unit 30 extracts a point cloud consisting of a point in the xy plane where the biological determination index I(x,y,t) shows a local peak at a specific time, and a plurality of points around that point (step SB51). Points showing local peaks can be extracted using, for example, a known CFAR algorithm. Alternatively, a point cloud consisting of a point where the power of the complex signal (x,y) shows a local peak, and a plurality of points around that point, may be extracted. In this case, points showing local peaks in a region determined to be a stationary object by the biological determination index I(x,y,t) are not extracted.

[0059] Next, the procedure for step SB51 will be explained with reference to Figure 12. Figure 12 is a schematic diagram showing an example of a point cloud 40 distributed in the observation range 50, i.e., the floor surface (xy plane). First, points 41P where the power of the complex signal P(x,y) with signals from stationary objects removed shows a local peak are extracted. In Figure 12, points 41P showing a local peak are represented by black-filled circles. In the example shown in Figure 12, four points 41P showing a local peak are extracted. For example, constant false alarm probability (CFAR) processing can be used to extract points 41P showing a local peak.

[0060] Once a point 41P indicating a local peak is extracted, several points 41A surrounding point 41P are extracted. In Figure 12, the surrounding points 41A are indicated by white circles. Various methods can be applied to extract the surrounding points 41A. For example, points where the power of the complex signal P(x,y) exceeds the threshold obtained by the CFAR process can be extracted as surrounding points 41A.

[0061] In addition, points 41A can be extracted as surrounding points 41A if their distance from point 41P, which exhibits a local peak, is less than a certain value. Furthermore, points 41A can also be extracted as surrounding points 41A if their signal-to-noise ratio (SNR) obtained from the complex signal P(x,y) is greater than or equal to a certain value.

[0062] After extracting the point cloud 40 (Figure 12) in step SB51 (Figure 11), clustering is performed on the point cloud 40 (step SB52). For example, clustering is performed based on the x and y coordinates of multiple points 41P, 41A included in the point cloud 40 extracted in step SB51. For clustering, well-known algorithms such as DBSCAN, k-means, x-means, etc., can be used.

[0063] Clustering is performed to classify the point cloud 40 into one or more clusters. In Figure 12, the point cloud is classified into three clusters 45A, 45B, and 45C. In the example shown in Figure 12, clusters 45A and 45B each contain one point 41P representing a local peak and multiple points 41A surrounding it. Cluster 45C contains two points 41P representing local peaks and multiple points 41A surrounding them.

[0064] After performing clustering in step S52 (Figure 11), the number of organisms in each cluster is determined based on the features of each cluster (step SB53). It is preferable to use features that have different values ​​depending on the number of organisms present in the range where the points included in the cluster are distributed (hereinafter sometimes simply referred to as the number of organisms in the cluster) as the features of the cluster. Next, the features of the clusters will be explained with reference to the diagrams from Figure 13A to Figure 13E.

[0065] Figures 13A to 13E are graphs showing the time evolution of various cluster features. The features shown in these graphs were calculated based on the complex signal P obtained when an evaluation experiment was conducted to actually detect living organisms. The horizontal axis represents the serial number assigned to the chirp frame, and the vertical axis represents the feature value. In other words, the horizontal axis corresponds to the passage of time. The solid lines in the graphs from Figures 13A to 13E show the features of clusters with 2 living organisms, and the dashed lines show the features of clusters with 1 living organism.

[0066] Figure 13A shows the time evolution of the feature when the total power of the complex signal P is used as the feature. The vertical axis represents the normalized value of the total power of the points included in the cluster. The total power of the cluster with 2 organisms is greater than the total power of the cluster with 1 organism. This result suggests that as the number of organisms in each of the clusters 45A, 45B, and 45C increases, the total power of each of the clusters 45A, 45B, and 45C also increases.

[0067] Figure 13B shows the time evolution of the feature when the average power of the complex signal P is used as the feature. The vertical axis represents the normalized value of the average power of all points included in the cluster. The average power of the cluster with 2 organisms is greater than the average power of the cluster with 1 organism. This result suggests that as the number of organisms in each of the clusters 45A, 45B, and 45C increases, the total power of each of the clusters 45A, 45B, and 45C also increases.

[0068] Figure 13C shows the time evolution of the feature when the number of points in a cluster is used as the feature. The vertical axis represents the normalized value of the number of points in a cluster. The number of points in a cluster with 2 organisms is greater than the number of points in a cluster with 1 organism. This result suggests that as the number of organisms in each of the clusters 45A, 45B, and 45C increases, the number of points constituting each of the clusters 45A, 45B, and 45C also increases.

[0069] Figures 13D and 13E show the time evolution of features when the standard deviation of the x-coordinate and y-coordinate of points included in the cluster are used as features, respectively. The vertical axis represents the normalized values ​​of the standard deviation of the x-coordinate and y-coordinate of points included in the cluster. The standard deviation of the coordinates of points included in clusters with 2 organisms is greater than the standard deviation of the coordinates of points included in clusters with 1 organism. This result suggests that as the number of organisms in each of clusters 45A, 45B, and 45C increases, the standard deviation of the x-coordinate and y-coordinate of points included in each of clusters 45A, 45B, and 45C also increases.

[0070] As explained with reference to Figures 13A to 13E, the total power, average power, number of points, standard deviation (or variance) of the x-coordinate, and standard deviation (or variance) of the y-coordinate of a cluster depend on the number of organisms in the cluster. Therefore, by setting a threshold for the features and comparing the calculated features with the threshold, the number of organisms in the cluster can be estimated.

[0071] As explained with reference to Figures 13A to 13E, the features of a cluster fluctuate over time. To avoid the effects of these time-dependent fluctuations, it is preferable to use the average value of the features over a certain period of time to determine the number of organisms. It is also preferable to determine the number of organisms based on multiple features. By using multiple features, the accuracy of the determination can be improved.

[0072] After determining the number of living organisms in step SB53 (Figure 11), the processing unit 30 determines the location of living organisms for each of the clusters 45A, 45B, and 45C (Figure 12) (step SB54). For example, the processing unit 30 determines the location of living organisms in each of the clusters 45A, 45B, and 45C based on the distribution of multiple points contained in each of the clusters 45A, 45B, and 45C.

[0073] Next, various examples of methods for determining the position of a living organism will be described with reference to Figures 14A to 15.

[0074] Figure 14A is a schematic diagram showing an example of the arrangement of clusters 45A and 45B when the point cloud within the observation range 50 is classified into two clusters 45A and 45B in step SB52 (Figure 11). The observation range 50 is divided into multiple sections Di. The point cloud constituting one cluster 45A is distributed within one section Da. The point cloud constituting the other cluster 45B is distributed across two sections Db and Dc. The centroids Ga and Gb of the point clouds constituting clusters 45A and 45B are located within sections Da and Db, respectively. In step SB53 (Figure 11), it is determined that the number of living organisms present in clusters 45A and 45B shown in Figure 14A is 1 and 2, respectively.

[0075] Figure 14B is a schematic diagram illustrating an example of a method for determining the location of living organisms. In the example shown in Figure 14B, the processing unit 30 determines that the number of living organisms in the cluster is present in the section Di which contains the centroids of the point clouds constituting the cluster. In the example shown in Figure 14A, since the centroid Ga of the point cloud of cluster 45A is located within section Da, the processing unit 30 determines that the number of living organisms in section Da is 1, which is the number of living organisms in cluster 45A. Since the centroid Gb of the point cloud of cluster 45B is located within section Db, the processing unit 30 determines that the number of living organisms in section Db is 2, which is the number of living organisms in cluster 45B. In this case, some points of the point cloud of cluster 45B are also distributed within section Dc, but the number of living organisms in section Dc is determined to be 0.

[0076] Figure 14C is a schematic diagram illustrating another example of a method for determining the location of living organisms. In the example shown in Figure 14C, it is determined that living organisms are distributed in one or more sections Di where the point cloud constituting the cluster is distributed. In the example shown in Figure 14A, since the point cloud constituting cluster 45A is distributed only within section Da, the processing unit 30 determines that the number of living organisms in section Da is 1, which is the number of living organisms in cluster 45A. Since the point cloud constituting cluster 45B is distributed across two sections Db and Dc, the processing unit 30 determines that living organisms exist in sections Db and Dc. Furthermore, since the number of living organisms in cluster 45B is 2, it is determined that the number of living organisms in sections Db and Dc is 1 each.

[0077] If a point cloud constituting a cluster with one organism is distributed across multiple partitions Di, it is advisable to determine one of these partitions Di as the partition where the organism resides. For example, the partition Di where the centroid of the point cloud is located, or the partition Di containing the most points, can be determined as the partition where the organism resides.

[0078] If the number of organisms in a cluster is greater than the number of partitions Di in which the point cloud constituting the cluster is distributed, it is advisable to allocate two or more organisms to one of the partitions Di. The allocation of organisms can be weighted, for example, by the number of points present in each partition Di.

[0079] If the number of organisms in a cluster is less than the number of compartments Di in which the point cloud constituting the cluster is distributed, it is advisable to set the number of adults in any of the compartments Di to 0. For example, the number of adults in the compartments Di should be set to 0 in order from the smallest number of points in each compartment Di to the largest.

[0080] Figure 15 is a schematic diagram illustrating yet another example of a method for determining the location of a living organism. In the example shown in Figure 15, the observation range 50 is not divided into sections, and the location of the living organism is defined by coordinates. For example, the processing unit 30 determines that a living organism exists at the centroid of the point cloud constituting the cluster. For example, the processing unit 30 determines that a living organism exists at the centroid Ga of the point cloud constituting cluster 45A, and determines the number of living organisms to be 1, which is the number of living organisms in cluster 45A. Also, the processing unit 30 determines that a living organism exists at the centroid Gb of the point cloud constituting cluster 45B, and determines the number of living organisms to be 2, which is the number of living organisms in cluster 45B.

[0081] Next, the excellent effects of the third embodiment will be described. In the third embodiment, as in the first embodiment, the position and number of living organisms can be determined while excluding stationary objects.

[0082] [Fourth Embodiment] Next, a biological detection device according to the fourth embodiment will be described with reference to Figure 16. Hereinafter, the configuration common to the biological detection device according to the third embodiment, described with reference to Figures 9 to 15, will be omitted from the explanation.

[0083] In the third embodiment (Figure 9), the amplitude component m(x,y) and phase component p(x,y) obtained from the three-dimensional data X(r,d,n) are two-dimensional data defined in the xy plane. In contrast, in the fourth embodiment, the amplitude component m(x,y,z) and phase component p(x,y,z) are three-dimensional data defined in three-dimensional space.

[0084] Figure 16 is a schematic diagram showing the arrangement of various pieces of information (data) that appear in the procedure for obtaining the amplitude component m(x, y, z) and phase component p(x, y, z), which are three-dimensional data, from the three-dimensional data X(r, d, n). Similar to the third embodiment shown in Figure 9, a complex signal P(r, θ, φ) is calculated from the three-dimensional data X(r, d, n). In the third embodiment (Figure 9), a two-dimensional complex signal P(x, y) is obtained from the three-dimensional complex signal P(r, θ, φ). In contrast, in the fourth embodiment, without obtaining two-dimensional data, the complex signal P(r, θ, φ) is converted into a complex signal P(x, y, z) defined in xyz orthogonal coordinates. Then, the amplitude component m(x, y, z) and phase component p(x, y, z) are calculated based on the complex signal P(x, y, z).

[0085] By calculating the amplitude component m(x, y, z) and phase component p(x, y, z) for multiple chirp frames, the discretization time t １ ,t ２ ,t ３ For each of the above, the amplitude component m(x, y, z) and the phase component p(x, y, z) are calculated.

[0086] In step S4 (Figure 10), a biological identification index is calculated for each position in the xyz three-dimensional space. Then, in step SB5 (Figure 10), the position of the living organism is identified in the three-dimensional space and the number of living organisms is counted. At this time, in each step from step SB51 to step SB54 (Figure 11), the two-dimensional space targeted in the third embodiment can be extended to a three-dimensional space and the same calculations can be performed.

[0087] Next, the excellent effects of the fourth embodiment will be described. In the fourth embodiment, the position and number of living organisms can be identified in three-dimensional space.

[0088] [Fifth Embodiment] Next, a biological detection device according to the fifth embodiment will be described with reference to Figures 17 and 18. Hereinafter, the description of components common to the biological detection device according to the third embodiment, which was described with reference to Figures 9 to 15, will be omitted. In the third embodiment, a biological determination index I defined by equation (3) is used to distinguish between living organisms and stationary objects, and living organisms are extracted. In the fifth embodiment, a biological determination index I' different from biological determination index I is used to extract living organisms. Furthermore, a complex signal P' originating from a living organism (a complex identifier originating from a living organism) is calculated from the biological determination index I'. The procedure for calculating the biological determination index I' and the complex signal P' will be described below.

[0089] Figure 17 is a schematic diagram showing the arrangement of various pieces of information (data) that appear in the procedure for obtaining the biological complex signal P'(r, θ, φ) from three-dimensional data X(r, d, n). First, the processing unit 30 obtains two-dimensional data X'(r, n) defined in a two-dimensional space of distance r and channel n by averaging the three-dimensional data X in the velocity direction (d direction). This procedure is the same as the procedure in the first embodiment (Figure 3).

[0090] Next, the two-dimensional data X'(r,n) is duplicated in the velocity direction (d direction) to calculate the three-dimensional data X'(r,d,n). The number of data points in the velocity direction is the same as the number of data points in the velocity direction in the three-dimensional data X(r,d,n).

[0091] Next, the biological detection index I'(r,d,n) is calculated by subtracting the three-dimensional data X'(r,d,n), which has been averaged in the velocity direction, from the original three-dimensional data X(r,d,n). Since the biological detection index I'(r,d,n) is obtained by subtracting the average velocity from each element of the three-dimensional data X(r,d,n), information caused by stationary objects that do not have velocity is removed, and information caused by living organisms, such as respiration, heartbeat, and body movement, which cause changes in the distance from the biological detection device 55 (Figure 6A) to the living organism surface, remains. Next, a complex signal caused by living organisms (biologically-induced complex signal) P'(r,θ,φ) is calculated from the biological detection index I'.

[0092] Next, with reference to Figure 18, the procedure for calculating the biologically-derived complex signal P' from the biological determination index I' will be explained.

[0093] Figure 18 is a schematic diagram showing the arrangement of various pieces of information (data) that appear in the procedure for obtaining the biologically-derived complex signal P'(r, θ, φ) from the biological judgment index I'(r, d, n).

[0094] From the biometric index I'(r, d, n), extract the i-th element in the r direction and the j-th element in the d direction. This will result in a one-dimensional biometric index I'(r, d, n) arranged in the channel direction (n direction). ｉ d ｊ ,n) is obtained. Biological judgment index I' (r ｉ d ｊ From n) the complex conjugate transpose matrix I' Ｈ (r ｉ d ｊ Find I' Ｈ (r ｉ d ｊ ,n) and the biological judgment index I'(r ｉ d ｊ From the matrix product with n), the correlation matrix R ＸＸ Calculate (n, n).

[0095] Next, by performing the same calculation for all elements in the r direction, we obtain an n × n × r three-dimensional correlation matrix R for each element in the d direction. ＸＸ We find (n, n, r). This gives us d three-dimensional correlation matrices R ＸＸ (n, n, r) is obtained. d three-dimensional correlation matrices R ＸＸ By averaging (n, n, r) in the d direction, a three-dimensional correlation matrix R is obtained. ＸＸ Find (n, n, r). Correlation matrix R ＸＸ By performing angle estimation processing using (n, n, r), the biologically-derived complex signal P'(r, θ, φ) is obtained.

[0096] Next, the biologically-derived complex signal P'(r,θ,φ) is transformed into an xyz Cartesian coordinate system to obtain the biologically-derived complex signal P'(x,y,z). Then, similar to step SB1 (Figure 11), a point cloud consisting of the point where the biologically-derived complex signal P'(x,y,z) shows a local peak, and multiple points surrounding it, is extracted. Next, the point cloud distributed in three dimensions is viewed from the z direction to transform it into a point cloud distributed in the xy plane.

[0097] Next, the location and number of living organisms are determined by performing the same procedure as in steps SB52 to SB54 in Figure 11.

[0098] Alternatively, before converting a point cloud distributed in three dimensions into a point cloud distributed in the xy plane, clustering can be performed on the point cloud in three-dimensional space, and then viewed from the z direction to convert it into a point cloud and clusters distributed in the xy plane.

[0099] Next, the superior effects of the fifth embodiment will be described. In the third embodiment (Figure 9), the biological determination index I is calculated using equations (1), (2), and (3) based on the amplitude component m(x,y) and phase component p(x,y) of the complex signal P(x,y), and signals from stationary objects other than living organisms are removed. In contrast, as in the fifth embodiment, it is also possible to remove signals from stationary objects by subtracting the three-dimensional data X'(r,d,n), which is averaged in the velocity direction, from the three-dimensional data X(r,d,n). By removing signals from stationary objects, the position and number of living organisms can be determined.

[0100] The embodiments described above are illustrative, and it goes without saying that partial substitution or combination of the configurations shown in different embodiments is possible. Similar effects and benefits from similar configurations in multiple embodiments will not be mentioned sequentially for each embodiment. Furthermore, the present invention is not limited to the embodiments described above. For example, it will be obvious to those skilled in the art that various modifications, improvements, and combinations are possible.

[0101] 20 Transmitting / receiving unit 21 Signal processing unit 22 Signal generation unit 23 AD converter 24 Mixer 30 Processing unit 40 Point cloud 41P Point showing a peak 41A Surrounding points 45A, 45B, 45C Cluster 50 Observation range 51 Object 52 Stationary object 52A Region corresponding to a stationary object 53 Person 53A Region corresponding to a person 54 Wall 54A Region corresponding to a wall 55 Biological detection device D1, D2, D3, D4, Da, Db, Dc, Di Diarranged section Rx Receiving antenna Tx Transmitting antenna

Claims

1. A biological detection device comprising: a transmitting and receiving unit that receives reflected waves of radio waves transmitted from at least one transmitting antenna to an observation range using multiple receiving antennas and generates a transmitted signal and an intermediate frequency signal based on the received signal; and a processing unit that identifies the location of a living organism within the observation range and counts the number of living organisms based on the intermediate frequency signal, wherein the processing unit calculates a complex signal for each location within the observation range that reflects the presence or absence of an object and the velocity of the object, based on the intermediate frequency signal; calculates a biological determination index for each location within the observation range that differs in size depending on whether the detected object is a living organism or not, and whose size depends on the number of living organisms, based on the complex signal; and identifies the location of a living organism within the observation range and counts the number of living organisms using the biological determination index.

2. The biological detection device according to claim 1, wherein the processing unit identifies the location of a living organism within the observation range and counts the number of living organisms, based on the distribution of the biological determination index for each location within the observation range, calculates the sum of the biological determination index in at least a portion of the area within the observation range, and counts the number of living organisms based on the sum of the biological determination index.

3. The biological detection device according to claim 2, wherein the processing unit divides the observation range into a plurality of sections, and when correcting the number of biological organisms, calculates the sum of the biological organism determination indices for each of the plurality of sections, and counts the number of biological organisms based on the sum of the biological organism determination indices as well.

4. The biological detection device according to claim 1, wherein the processing unit, in the process of identifying the location of a living organism within the observation range and counting the number of living organisms, extracts a point cloud consisting of a point within the observation range where either the absolute value of the complex signal or the biological determination index shows a local peak, and a plurality of points around the point showing the local peak, and identifies the location of a living organism and counts the number of living organisms based on the extracted point cloud.

5. The biological detection device according to claim 4, wherein the processing unit performs clustering on the extracted point cloud to classify the point cloud into one or more clusters containing multiple points, identifies the location of a biological organism based on the cluster, and counts the number of biological organisms.

6. The biological detection device according to claim 5, wherein the processing unit determines the number of biological organisms corresponding to the clusters based on the respective feature quantities of the clusters.

7. The biological detection device according to claim 5 or 6, wherein the processing unit determines the position of a living organism based on the distribution of a plurality of points included in each of the clusters.

8. The biological detection device according to any one of claims 1 to 7, wherein the processing unit uses an index that includes the strength of the correlation between the time change of the amplitude component of the complex signal and the time change of the phase component of the complex signal for each position within the observation range as the biological determination index.

9. The biodetector according to any one of claims 1 to 8, wherein the transmitted signal is a frequency continuous modulated wave in which chirps whose frequency changes linearly with respect to time repeatedly appear, and the processing unit performs a distance FFT for each chirp, performs a velocity FFT on the result of the distance FFT for each chirp frame containing a plurality of chirps, and calculates the complex signal for each position within the observation range.

Images