Signal processing system, human detection system, signal processing method and program

Abstract

The present disclosure addresses the problem of suppressing deterioration in the accuracy of determining the presence of people even when the determination is performed in a plurality of detection areas by using human biometric information. A signal processing system (2) comprises a signal acquisition unit (231), a waveform estimation unit (232), a waveform comparison unit (233), and a determination unit (234). The signal acquisition unit (231) acquires, from a sensor system (10), a plurality of detection signals (for example, detection signals L1, L2) respectively corresponding to a plurality of detection areas. The sensor system (10) is provided in a space, and detects people in a plurality of different detection areas. The waveform estimation unit (232) estimates a biometric waveform relating to human biometric information from each of the plurality of detection signals. The waveform estimation unit (232) compares the plurality of biometric waveforms respectively corresponding to the plurality of detection signals. The determination unit (234) determines presence of people in the space on the basis of the result of comparison by the waveform comparison unit (233).

Description

[Technical Field]

[0001] This disclosure generally relates to signal processing systems, human detection systems, signal processing methods, and programs, and more specifically to signal processing systems, human detection systems, signal processing methods, and programs for determining the presence of a person in space. [Background technology]

[0002] Conventionally, sensor systems that use radio wave sensors to detect the state of a person within a predetermined area, such as a person's movement or presence, are known (see, for example, Patent Document 1).

[0003] The radio wave sensor in the sensor system of Patent Document 1 outputs radio waves within a detection area, receives radio waves reflected by objects within the detection area, and outputs a radio wave sensor signal corresponding to the state of the object. The signal processing system in the sensor system of Patent Document 1 generates time-series data of human biological information by applying signal processing to the radio wave sensor signal, and uses the time-series data to determine the state of the person.

[0004] It is conceivable to install multiple radio wave sensors in multiple detection areas, each distinct from the others, and to determine the state of a person in each detection area. In this case, a radio wave sensor targeting one of the multiple detection areas may receive radio waves reflected from other detection areas. Since the sensor system in Patent Document 1 does not take into account the reception of radio waves reflected from other detection areas, applying the sensor system in Patent Document 1 may reduce the accuracy of determining the state (presence) of a person. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Japanese Patent Publication No. 2021-171271 [Overview of the Initiative]

[0006] This disclosure is made in view of the above issues and aims to provide a signal processing system, a human detection system, a signal processing method, and a program that can suppress a decrease in the accuracy of the determination even when determining the presence of a person in multiple detection areas using human biometric information.

[0007] A signal processing system according to one aspect of the present disclosure comprises a signal acquisition unit, a waveform estimation unit, a waveform comparison unit, and a determination unit. The signal acquisition unit acquires a plurality of detection signals corresponding to each of the plurality of detection areas from a sensor system installed in space and detecting people in a plurality of different detection areas. The aforementioned plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. The waveform estimation unit estimates a bio-waveform relating to a person's biological information from each of the plurality of detection signals. The waveform comparison unit compares the plurality of bio-waveforms corresponding to each of the plurality of detection signals. The determination unit determines, based on the comparison result of the waveform comparison unit, the human in the space. number The waveform comparison unit calculates an evaluation index representing the similarity of the plurality of biological waveforms as a result of the comparison. Based on the evaluation index, the determination unit determines the human in the space. number The evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of the plurality of biological waveforms over a predetermined period. The waveform comparison unit uses a cross-correlation function to calculate a plurality of cross-correlation coefficients for the plurality of biological waveforms over the predetermined period as the comparison result. The determination unit determines the value of a person in the space based on the cross-correlation coefficients calculated by the waveform comparison unit. number The waveform comparison unit calculates multiple cross-correlation coefficients for each of the multiple predetermined periods. The waveform comparison unit extracts the cross-correlation coefficient corresponding to the peak value in the frequency distribution of the multiple cross-correlation coefficients as the maximum cross-correlation coefficient. The waveform comparison unit uses the calculated value obtained from the extracted maximum cross-correlation coefficients as the comparison result. Based on the calculated value, the determination unit determines whether a person in the space is number Determine. A signal processing system according to one aspect of the present disclosure comprises a signal acquisition unit, a waveform estimation unit, a waveform comparison unit, and a determination unit. The signal acquisition unit acquires a plurality of detection signals corresponding to each of the plurality of detection areas from a sensor system installed in space and detecting people in a plurality of different detection areas. The aforementioned plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. The waveform estimation unit estimates a bio-waveform relating to a person's biological information from each of the plurality of detection signals. The waveform comparison unit compares the plurality of bio-waveforms corresponding to each of the plurality of detection signals. The determination unit determines, based on the comparison result of the waveform comparison unit, the human in the space. number The waveform comparison unit calculates an evaluation index representing the similarity of the plurality of biological waveforms as a result of the comparison. Based on the evaluation index, the determination unit determines the human in the space. number The evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of the plurality of biological waveforms over a predetermined period. The waveform comparison unit uses a cross-correlation function to calculate a plurality of cross-correlation coefficients for the plurality of biological waveforms over the predetermined period as the comparison result. The determination unit determines the value of a person in the space based on the cross-correlation coefficients calculated by the waveform comparison unit. number The determination is made. The plurality of detection signals include a first detection signal and a second detection signal. The waveform estimation unit estimates a first biological waveform corresponding to the first detection signal and a second biological waveform corresponding to the second detection signal. The waveform comparison unit calculates a plurality of cross-correlation coefficients for the first biological waveform and the second biological waveform over the predetermined period as the comparison result. The determination unit determines whether there is one person or two people in the space based on the comparison result and a predetermined threshold. The predetermined threshold is a value based on the reference cross-correlation coefficient corresponding to the peak value in the frequency distribution of the plurality of reference cross-correlation coefficients, which are calculated using the first person biological waveform estimated by the waveform estimation unit when only the first person is present in the space and the second person biological waveform estimated by the waveform estimation unit when only the second person is present in the space over the predetermined period. A signal processing system according to one aspect of the present disclosure comprises a signal acquisition unit, a waveform estimation unit, a waveform comparison unit, and a determination unit. The signal acquisition unit acquires a plurality of detection signals corresponding to each of the plurality of detection areas from a sensor system installed in space and detecting people in a plurality of different detection areas. The aforementioned plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. The waveform estimation unit estimates a bio-waveform relating to a person's biological information from each of the plurality of detection signals. The waveform comparison unit compares the plurality of bio-waveforms corresponding to each of the plurality of detection signals. The determination unit determines, based on the comparison result of the waveform comparison unit, the human in the space. number The waveform comparison unit calculates an evaluation index representing the similarity of the plurality of biological waveforms as a result of the comparison. Based on the evaluation index, the determination unit determines the human in the space. number The evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of the plurality of biological waveforms over a predetermined period. The waveform comparison unit uses a cross-correlation function to calculate a plurality of cross-correlation coefficients for the plurality of biological waveforms over the predetermined period as the comparison result. The determination unit determines the value of a person in the space based on the cross-correlation coefficients calculated by the waveform comparison unit. numberThe determination is made. The plurality of detection signals include a first detection signal and a second detection signal. The waveform estimation unit estimates a first biological waveform corresponding to the first detection signal and a second biological waveform corresponding to the second detection signal. The waveform comparison unit calculates a plurality of cross-correlation coefficients for the first biological waveform and the second biological waveform over the predetermined period as the comparison result. The determination unit determines whether there is one person or two people in the space based on the comparison result and a predetermined threshold. The waveform estimation unit includes a first waveform estimation unit that estimates the first biological waveform and a second waveform estimation unit that estimates the second biological waveform. The predetermined threshold is a value based on a reference cross-correlation coefficient corresponding to the peak value in a frequency distribution which is a combination of a plurality of first reference cross-correlation coefficients, which are a plurality of reference cross-correlation coefficients over the predetermined period when only the first person is present in the space, and a plurality of second reference cross-correlation coefficients, which are a plurality of reference cross-correlation coefficients over the predetermined period when only the second person is present in the space. The plurality of first reference cross-correlation coefficients are obtained from the first waveform as the first biological waveform for the predetermined period estimated by the first waveform estimation unit and the second waveform as the second biological waveform for the predetermined period estimated by the second waveform estimation unit, when only the first person is present in the space. The plurality of second reference cross-correlation coefficients are obtained from the third waveform as the first biological waveform for the predetermined period estimated by the first waveform estimation unit and the fourth waveform as the second biological waveform for the predetermined period estimated by the second waveform estimation unit, when only the second person is present in the space. A signal processing system according to one aspect of the present disclosure comprises a signal acquisition unit, a waveform estimation unit, a waveform comparison unit, a determination unit, a waveform adjustment unit, and a display processing unit. The signal acquisition unit acquires a plurality of detection signals corresponding to each of the plurality of detection areas from a sensor system installed in space and detecting people in a plurality of different detection areas. The aforementioned plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. The waveform estimation unit estimates a bio-waveform relating to a person's biological information from each of the plurality of detection signals. The waveform comparison unit compares the plurality of bio-waveforms corresponding to each of the plurality of detection signals. The determination unit determines, based on the comparison result of the waveform comparison unit, the human in the space. numberThe waveform comparison unit calculates an evaluation index representing the similarity of the plurality of biological waveforms as a result of the comparison. Based on the evaluation index, the determination unit determines the human in the space. number The evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of the plurality of biological waveforms over a predetermined period. The waveform comparison unit uses a cross-correlation function to calculate a plurality of cross-correlation coefficients for the plurality of biological waveforms over the predetermined period as the comparison result. The determination unit determines the value of a person in the space based on the cross-correlation coefficients calculated by the waveform comparison unit. number The waveform estimation unit determines whether there is one person or two people in the space. The plurality of detection signals include a first detection signal and a second detection signal. The waveform estimation unit estimates a first biological waveform corresponding to the first detection signal and a second biological waveform corresponding to the second detection signal. The waveform comparison unit calculates a plurality of cross-correlation coefficients for the first biological waveform and the second biological waveform over a predetermined period as the comparison result. The determination unit determines whether there is one person or two people in the space based on the comparison result and a predetermined threshold. If the determination unit determines that there is one person in the space, the waveform adjustment unit calculates the first average amplitude of the first biological waveform and the second average amplitude of the second biological waveform, and adjusts the amplitude of the biological waveform with a smaller average amplitude by multiplying it by a suppression coefficient, while not adjusting the amplitude of the biological waveform with a larger average amplitude. If the determination unit determines that there are two people in the space, the waveform adjustment unit does not adjust the amplitude of either the first biological waveform or the second biological waveform. If the determination unit determines that there is one person in the space, the display processing unit displays the biological waveform whose amplitude has been adjusted from the first biological waveform and the second biological waveform, as well as the biological waveform with a large average amplitude that has not been adjusted from the first biological waveform and the second biological waveform. If the determination unit determines that there are two people in the space, the display processing unit displays the first biological waveform and the second biological waveform that have not been adjusted in amplitude on the display processing unit.

[0008] A human detection system according to one aspect of the present disclosure includes the signal processing system and a sensor system that detects a person in the space.

[0009] A signal processing method according to one aspect of the present disclosure includes a signal acquisition step, a waveform estimation step, a waveform comparison step, and a determination step. The signal acquisition step acquires a plurality of detection signals respectively corresponding to a plurality of detection areas from a sensor system that is provided in a space and detects a person in a plurality of different detection areas. The aforementioned plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. The waveform estimation step estimates a biological waveform related to a person's biological information from each of the plurality of detection signals. The waveform comparison step compares the plurality of biological waveforms respectively corresponding to the plurality of detection signals. The determination step determines the presence or absence of a person in the space based on the comparison result in the waveform comparison step. In the waveform comparison step, an evaluation index representing the similarity of the plurality of biological waveforms is calculated as the comparison result. In the determination step, the presence or absence of a person in the space is determined based on the evaluation index. number The evaluation index is a correlation coefficient representing the mutual correlation relationship of the plurality of biological waveforms in a predetermined period. In the waveform comparison step, a plurality of the correlation coefficients for the plurality of biological waveforms in the predetermined period are calculated as the comparison result using a correlation function. In the determination step, the presence or absence of a person in the space is determined based on the correlation coefficient calculated in the waveform comparison step. number In the waveform comparison step, the correlation coefficients in the plurality of correlation coefficients in the predetermined period are calculated for each of the plurality of predetermined periods. In the waveform comparison step, the correlation coefficient corresponding to the peak value in the frequency distribution of the plurality of correlation coefficients among the plurality of correlation coefficients is extracted as the maximum correlation coefficient. In the waveform comparison step, an arithmetic value obtained from the plurality of extracted maximum correlation coefficients is used as the comparison result. In the determination step, the presence or absence of a person in the space is determined based on the arithmetic value. number The presence or absence of a person in the space is determined. number The presence or absence of a person in the space is determined.

[0010] A program according to one aspect of the present disclosure is a program for causing a computer system to execute the signal processing method.

Brief Description of Drawings

[0011] [Figure 1] FIG. 1 is a block diagram showing the configuration of a signal processing system according to Embodiment 1 of the present disclosure and a human detection system including the same signal processing system. [Figure 2] FIG. 2 is a system configuration diagram showing an example of use of the human detection system described above. [Figure 3] FIGS. 3A and 3B are waveform diagrams showing an example of a biological waveform estimated by the signal processing system described above. [Figure 4] FIG. 4 is a diagram showing the relationship between the cross-correlation coefficient obtained from the waveform diagram described above and time (seconds). [Figure 5] FIG. 5A is a waveform diagram showing an example of a first biological waveform and a second biological waveform. FIG. 5B is a transition diagram showing an example of the transition of the cross-correlation coefficient obtained from the first biological waveform and the second biological waveform. FIG. 5C is a transition diagram showing an example of the transition of the determination result. [Figure 6] FIG. 6A is a waveform diagram showing another example of a first biological waveform and a second biological waveform. FIG. 6B is a transition diagram showing another example of the transition of the cross-correlation coefficient obtained from the first biological waveform and the second biological waveform. FIG. 6C is a transition diagram showing another example of the transition of the determination result. [Figure 7] FIGS. 7A and 7B are diagrams showing the relationship between the comparison result and a predetermined threshold value serving as a determination criterion. [Figure 8] FIG. 8 is a flowchart showing the operation of the signal processing system described above. [Figure 9] FIG. 9 is a block diagram showing the configuration of a signal processing system according to Modification 1 of Embodiment 1 and a human detection system including the same signal processing system. [Figure 10] FIG. 10 is a system configuration diagram showing an example of use of the human detection system described above. [Figure 11]Figure 11 is a block diagram showing the configuration of a signal processing system according to a modified example 2 of Embodiment 1, and a human detection system equipped with the same signal processing system. [Figure 12] Figure 12 shows the relationship between the cross-correlation coefficient and the probability that there are two people in a space. [Figure 13] Figure 13 is a block diagram showing the configuration of a signal processing system according to Embodiment 2 of the present disclosure, and a human detection system equipped with the above signal processing system. [Figure 14] Figure 14 shows the relationship between the cross-correlation coefficient and the suppression coefficient. [Figure 15] Figures 15A and 15B are waveform diagrams showing the biological waveform before and after processing by the waveform adjustment unit of the signal processing system described above. [Figure 16] Figure 16 is a flowchart illustrating the operation of the signal processing system described above. [Figure 17] Figure 17 is a flowchart showing the operation of a signal processing system according to a modified example 1 of Embodiment 2. [Modes for carrying out the invention]

[0012] The embodiments and modifications described below are merely examples of the disclosure, and the disclosure is not limited to these embodiments and modifications. Various modifications are possible, including design changes, as long as they do not depart from the technical concept of the disclosure.

[0013] (Embodiment 1) The human detection system 1 and the signal processing system provided for the human detection system 1 according to Embodiment 1 will be described below with reference to Figures 1 to 8.

[0014] (1) Overview As shown in Figure 1, the human detection system 1 according to Embodiment 1 comprises a sensor system 10 and a signal processing device 20 as a signal processing system 2.

[0015] The sensor system 10 is installed in space SP1 and detects people (for example, people u1 and u2 shown in Figure 2) in multiple distinct detection areas G1 and G2. The sensor system 10 outputs radio waves into the multiple detection areas G1 and G2 in space SP1, receives radio waves reflected by objects in the multiple detection areas G1 and G2, and outputs a radio wave sensor signal according to the state of the object. The state of the object is such as the object's movement, presence or absence, speed of movement, or position. In this disclosure, people u1 and u2 are assumed to be objects, and the state of people u1 and u2 is biological information (respiration, heart rate, pulse, etc.). In particular, in this disclosure, the biological information is the person's respiration.

[0016] As shown in Figure 1, the signal processing device 20 (signal processing system 2) comprises a signal acquisition unit 231, a waveform estimation unit 232, a waveform comparison unit 233, and a determination unit 234. The signal acquisition unit 231 acquires multiple detection signals corresponding to multiple detection areas G1 and G2 from a sensor system 10 installed in space SP1 that detects people in multiple different detection areas G1 and G2. The waveform estimation unit 232 estimates bio-waveforms relating to the bio-information of people u1 and u2 from each of the multiple detection signals. The waveform comparison unit 233 compares multiple bio-waveforms corresponding to the multiple detection signals. The determination unit 234 determines the presence of people in space SP1 based on the comparison results of the waveform comparison unit 233.

[0017] This configuration makes it possible to suppress a decrease in the accuracy of detection even when using a person's biometric information to determine the presence of a person in multiple detection areas.

[0018] (2) Composition (2.1) Sensor System The sensor system 10 includes a plurality (two in the illustrated example) of radio wave sensors 11, as shown in Figure 1.

[0019] Multiple radio wave sensors 11 detect biometric information (human respiration) of people (for example, people u1 and u2 shown in Figure 2) in multiple different detection areas G1 and G2. The multiple radio wave sensors 11 are, for example, Doppler sensors, FMCW (Frequency-Modulated Continuous-Wave) type radio wave sensors, etc. When it is necessary to refer to the multiple radio wave sensors 11 individually, they are referred to as the first radio wave sensor 11a, the second radio wave sensor 11b, and so on.

[0020] As shown in Figure 1, the radio wave sensor 11 comprises a sensor unit 101, a signal generation unit 102, and a communication unit 103. Hereinafter, the sensor unit 101, signal generation unit 102, and communication unit 103 of the first radio wave sensor 11a may be referred to as sensor unit 101a, signal generation unit 102a, and communication unit 103a. Similarly, the sensor unit 101, signal generation unit 102, and communication unit 103 of the second radio wave sensor 11b may be referred to as sensor unit 101b, signal generation unit 102b, and communication unit 103b.

[0021] The sensor unit 101 outputs radio waves within the detection area in space SP1 and receives radio waves reflected by objects within the detection area. For example, the sensor unit 101a of the first radio wave sensor 11a outputs radio waves within the detection area G1 and receives radio waves reflected by objects within the detection area G1. The sensor unit 101b of the second radio wave sensor 11b outputs radio waves within the detection area G2 and receives radio waves reflected by objects within the detection area G2.

[0022] The signal generation unit 102 generates a detection signal based on an oscillation signal corresponding to the output radio wave and a reception signal corresponding to the received radio wave. For example, the detection signal is a Doppler signal that includes in-phase and quadrature components. The signal generation unit 102 generates a detection signal (Doppler signal) based on a signal that has a component of the frequency difference between the reception signal and the oscillation signal. For example, the signal generation unit 102a of the first radio wave sensor 11a generates a detection signal (Doppler signal) L1 based on an oscillation signal corresponding to the radio wave output by the sensor unit 101a and a reception signal corresponding to the radio wave received by the sensor unit 101a. The signal generation unit 102b of the second radio wave sensor 11b generates a detection signal (Doppler signal) L2 based on an oscillation signal corresponding to the radio wave output by the sensor unit 101b and a reception signal corresponding to the radio wave received by the sensor unit 101b. In other words, the multiple detection signals include the first detection signal and the second detection signal. Hereafter, detection signal L1 may be referred to as first detection signal L1, and detection signal L2 as second detection signal L2.

[0023] The communication unit 103 has a communication interface for communicating with the signal processing unit 20. The communication unit 103 outputs (transmits) the detection signal generated by the signal generation unit 102 to the signal processing unit 20. For example, the communication unit 103a of the first radio wave sensor 11a outputs the first detection signal L1 generated by the signal generation unit 102a to the signal processing unit 20. The communication unit 103b of the second radio wave sensor 11b transmits the second detection signal L2 generated by the signal generation unit 102b to the signal processing unit 20.

[0024] (2.2) Signal processing equipment (signal processing system) As shown in Figure 1, the signal processing device 20 includes a communication unit 21, a display unit 22, and a control unit 23.

[0025] The signal processing device 20 includes, for example, a computer system having a processor and memory. The computer system functions as a control unit 23 when the processor executes a program stored in memory. The program executed by the processor is pre-recorded in the computer system's memory, but it may also be provided on a recording medium such as a memory card, or provided via a telecommunications line such as the Internet.

[0026] The communication unit 21 has a communication interface for communicating with multiple radio wave sensors 11 of the sensor system 10. For example, the communication unit 21 receives a first detection signal L1 output from the first radio wave sensor 11a. The communication unit 21 also receives a second detection signal L2 output from the second radio wave sensor 11b.

[0027] The display unit 22 is a thin display device such as a liquid crystal display or an organic electroluminescent (EL) display. The display unit 22 displays a biological waveform based on the detection signal received from the sensor system 10. Here, the biological waveform is, for example, a human respiratory waveform.

[0028] As shown in Figure 1, the control unit 23 includes a signal acquisition unit 231, a waveform estimation unit 232, a waveform comparison unit 233, a determination unit 234, and a display processing unit 235.

[0029] The signal acquisition unit 231 acquires multiple detection signals corresponding to multiple detection areas G1 and G2 from a sensor system 10 installed in space SP1 that detects people in multiple different detection areas G1 and G2, via the communication unit 21. The signal acquisition unit 231 includes a first signal acquisition unit 241 and a second signal acquisition unit 242. The first signal acquisition unit 241 acquires a first detection signal L1 output from a first radio wave sensor 11a. The second signal acquisition unit 242 acquires a second detection signal L2 output from a second radio wave sensor 11b.

[0030] The waveform estimation unit 232 estimates a bio-waveform related to a person's biological information from each of the multiple detection signals acquired by the signal acquisition unit 231. In this case, the waveform estimation unit 232 estimates a person's respiratory waveform from each of the multiple detection signals acquired by the signal acquisition unit 231. The waveform estimation unit 232 estimates multiple bio-waveforms corresponding to each of the multiple detection signals using a data estimation learning model constructed by a recurrent neural network.

[0031] The waveform estimation unit 232 includes a first waveform estimation unit 251 and a second waveform estimation unit 252.

[0032] The first waveform estimation unit 251 estimates the biological waveform (respiratory waveform) based on the first detection signal L1 acquired by the first signal acquisition unit 241. The second waveform estimation unit 252 estimates the biological waveform (respiratory waveform) based on the second detection signal L2 acquired by the second signal acquisition unit 242. In other words, the waveform estimation unit 232 estimates the biological waveform corresponding to the first detection signal L1 (first biological waveform) and the biological waveform corresponding to the second detection signal L2 (second biological waveform).

[0033] The first waveform estimation unit 251 estimates a person's biological waveform from a first detection signal L1 using a learning model. The second waveform estimation unit 252 estimates a person's biological waveform from a second detection signal L2 using the same learning model. The learning model is a data estimation learning model constructed using a recurrent neural network (deep learning). The first waveform estimation unit 251 uses a recurrent neural network (RNN), a long short-term memory (LSTM), or a gated recurrent unit (GRU) as the recurrent neural network. The learning model is pre-generated by supervised learning using a large number of learning detection signals as learning data (training data). By using the learning model constructed by the pre-generated recurrent neural network described above, the first waveform estimation unit 251 and the second waveform estimation unit 252 can improve the estimation accuracy of biological waveforms that fluctuate over time. When training a neural network, it is preferable to use highly reliable bio-waveforms obtained from contact sensors directly attached to the human body as training data. Furthermore, it is preferable that the training model is adjusted so that when the quality of the input detection signal (the signal level itself or the signal-to-noise ratio) is good, the amplitude of the output bio-waveform is normalized to approximately 1, and when the quality of the detection signal is low or the signal is one that the training model cannot infer, the amplitude of the bio-waveform is adjusted to be smaller than 1.

[0034] The waveform comparison unit 233 compares multiple biological waveforms corresponding to multiple detection signals. Specifically, the waveform comparison unit 233 compares biological waveform W1 based on the first detection signal L1 (see Figure 3A) with biological waveform W2 based on the second detection signal L2 (see Figure 3B). Hereafter, biological waveform W1 may be referred to as the first biological waveform W1, and biological waveform W2 as the second biological waveform W2.

[0035] The waveform comparison unit 233 calculates an evaluation index representing the similarity of multiple biological waveforms (e.g., biological waveforms W1, W2) as a comparison result. For example, the evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of multiple biological waveforms (e.g., biological waveforms W1, W2) over a predetermined period. The waveform comparison unit 233 uses a cross-correlation function to calculate multiple cross-correlation coefficients for multiple biological waveforms over a predetermined period and extracts the calculation results as comparison results. In this embodiment, the waveform comparison unit 233 calculates multiple cross-correlation coefficients for the first biological waveform W1 and the second biological waveform W2 over a predetermined period as comparison results.

[0036] For example, the waveform comparison unit 233 uses the following equation 1 to determine the cross-correlation coefficient between the first biological waveform W1 and the second biological waveform W2 over time. Here, XC is the cross-correlation coefficient, corr() is the cross-correlation function, and max() is the maximum value.

[0037]

number

[0038] Specifically, the waveform comparison unit 233 calculates the cross-correlation coefficient as the maximum value obtained by normalizing the cross-correlation function between the first biological waveform W1 and the second biological waveform W2. The cross-correlation coefficient is a number that is greater than or equal to 0 and less than or equal to 1. Figure 4 shows the cross-correlation coefficient calculated by the waveform comparison unit 233 over time. The waveform Z1 shown in Figure 4 represents the change in the cross-correlation coefficient calculated over time. Figures 5A to 5C show an example (approximately 6 hours) of the waveform diagrams of the first and second biological waveforms, the time-series output of the cross-correlation coefficient, and the judgment result, for example, when sleeping for one night. Figures 6A to 6C show another example (approximately 6 hours) of the waveform diagrams of the first and second biological waveforms, the time-series output of the cross-correlation coefficient, and the judgment result, for example, when sleeping for one night. Figures 5A to 5C are examples where there is one person in space SP1, and Figures 6A to 6C are examples where there are two people in space SP1.

[0039] Figures 5A and 6A show the first biological waveforms W1a and W1b and the second biological waveforms W2a and W2b. Figure 5B shows waveform Z11, which represents the change in the cross-correlation coefficient calculated by the waveform comparison unit 233 from the first biological waveform W1a and the second biological waveform W2a. Figure 6B shows waveform Z12, which represents the change in the cross-correlation coefficient calculated by the waveform comparison unit 233 from the first biological waveform W1b and the second biological waveform W2b. Figure 5C shows waveform O11, which represents the change in the judgment result output by the judgment unit 234 based on waveform Z11, which represents the change in the cross-correlation coefficient. Figure 6C shows waveform O12, which represents the change in the judgment result output by the judgment unit 234 based on waveform Z12, which represents the change in the cross-correlation coefficient.

[0040] The determination unit 234 determines the presence of a person in space SP1 based on the comparison results of the waveform comparison unit 233. In other words, the determination unit 234 determines the presence of a person in space SP1 based on evaluation indicators. More specifically, the determination unit 234 determines the presence of a person in space SP1 based on the cross-correlation coefficient extracted by the waveform comparison unit 233.

[0041] Specifically, the determination unit 234 determines whether there is one person or two people in space SP1 based on the determination result and a predetermined threshold. The determination unit 234 determines that there is one person in space SP1 if the cross-correlation coefficient, which is the comparison result, is greater than or equal to the predetermined threshold. The determination unit 234 determines that there are two people in space SP1 if the cross-correlation coefficient, which is the comparison result, is less than the predetermined threshold.

[0042] Here, the predetermined threshold is a value based on the reference cross-correlation coefficient corresponding to the peak value in the frequency distribution of multiple reference cross-correlation coefficients calculated using the first and second human biological waveforms over a predetermined period. The first human biological waveform is the biological waveform estimated by the first waveform estimation unit 251 of the waveform estimation unit 232 when only the first person (e.g., person u1) is present in space SP1. The second human biological waveform is the biological waveform estimated by the second waveform estimation unit 252 of the waveform estimation unit 232 when only the second person (e.g., person u2) is present in space SP1. For example, the predetermined threshold is a value greater than the reference cross-correlation coefficient corresponding to the peak value in the frequency distribution of the calculated multiple reference cross-correlation coefficients. Alternatively, the predetermined threshold may be a value based on the reference cross-correlation coefficient corresponding to the median value in the frequency distribution of multiple reference cross-correlation coefficients calculated using the first and second human biological waveforms over a predetermined period.

[0043] For example, Figure 7A shows the frequency distribution of the cross-correlation coefficient when there is one person in space SP1, and Figure 7B shows the frequency distribution of the cross-correlation coefficient when there are two people in space SP1. Here, we consider TH1 as a predetermined threshold. As shown in Figure 7A, the comparison results of the waveform comparison unit 233 can be set so that most of them are distributed in a region greater than the predetermined threshold TH1, that is, the cross-correlation coefficient R1 corresponding to the peak value of the distribution is greater than or equal to the predetermined threshold TH1. Also, as shown in Figure 7B, the comparison results of the waveform comparison unit 233 can be set so that most of them are distributed in a region smaller than the predetermined threshold TH1, that is, the cross-correlation coefficient R2 corresponding to the peak value of the distribution is smaller than the predetermined threshold TH1. By setting the threshold TH1 in this way, if the comparison result of the waveform comparison unit 233 is greater than or equal to the predetermined threshold TH1, the determination unit 234 determines that there is one person in space SP1, and if the comparison result of the waveform comparison unit 233 is smaller than the predetermined threshold TH1, the determination unit 234 determines that there are two people in space SP1.

[0044] The determination performed by the determination unit 234 will be explained in detail using Figures 5B, 5C, 6B, and 6C. Let's assume a predetermined threshold is set to, for example, "0.5". In the example in Figure 5B, the comparison result output from the waveform comparison unit 233, i.e., waveform Z11, is compared with the predetermined threshold "0.5". Figure 5C shows the comparison result between waveform Z11 and the predetermined threshold "0.5", i.e., the determination result. In Figure 5C, the progress of the results is shown, where the comparison result output from the waveform comparison unit 233 is greater than or equal to the predetermined threshold "0.5", is "1" (corresponding to 1 person), and the rest is "2" (corresponding to 2 people). In Figure 5C, the majority of the determination results are "1", and it can be determined in the time series that at least one person is present during that time period. In the example in Figure 6B, the comparison result output from the waveform comparison unit 233, i.e., waveform Z12, is compared with the predetermined threshold "0.5". Figure 6C shows the comparison result between waveform Z12 and the predetermined threshold "0.5", i.e., the determination result. Figure 6C shows the progression of the results output from the waveform comparison unit 233, where the comparison result is greater than or equal to a predetermined threshold of "0.5" is output as "1" (corresponding to 1 person), and all other locations are output as "2" (corresponding to 2 people). In Figure 6C, the majority of the judgment results are "2," indicating that at least two people were present during that time period, as determined over time.

[0045] The display processing unit 235 displays the determination result from the determination unit 234 on the display unit 22. Furthermore, the display processing unit 235 displays the biological waveform estimated by the first waveform estimation unit 251 (first biological waveform) and the biological waveform estimated by the second waveform estimation unit 252 (second biological waveform) on the display unit 22.

[0046] (3) Operation Here, the operation of the signal processing device 20 will be explained using Figure 8.

[0047] The signal acquisition unit 231 acquires multiple detection signals from the sensor system 10, each corresponding to a plurality of detection areas G1 and G2 (step S1). Specifically, the first signal acquisition unit 241 of the signal acquisition unit 231 acquires the first detection signal L1 output from the first radio wave sensor 11a. The second signal acquisition unit 242 of the signal acquisition unit 231 acquires the second detection signal L2 output from the second radio wave sensor 11b.

[0048] The waveform estimation unit 232 performs waveform estimation processing (step S2). Specifically, the first waveform estimation unit 251 of the waveform estimation unit 232 estimates the first biological waveform W1 based on the first detection signal L1 acquired by the first signal acquisition unit 241. The second waveform estimation unit 252 of the waveform estimation unit 232 estimates the second biological waveform W2 based on the second detection signal L2 acquired by the second signal acquisition unit 242.

[0049] The waveform comparison unit 233 performs waveform comparison processing (step S3). Specifically, the waveform comparison unit 233 compares the first biological waveform W1 with the second biological waveform W2. More specifically, the waveform comparison unit 233 uses a cross-correlation function to calculate multiple cross-correlation coefficients for multiple biological waveforms over a predetermined period.

[0050] The determination unit 234 performs a determination process (step S4). The determination unit 234 determines the presence of a person in space SP1 based on the cross-correlation coefficient extracted as a comparison result by the waveform comparison unit 233. Specifically, the determination unit 234 determines that there is one person in space SP1 if the cross-correlation coefficient, which is the comparison result, is greater than or equal to a predetermined threshold. The determination unit 234 determines that there are two people in space SP1 if the cross-correlation coefficient, which is the comparison result, is less than a predetermined threshold.

[0051] The display processing unit 235 performs display processing (step S5). The display processing unit 235 displays the determination result from the determination unit 234, the biological waveform estimated by the first waveform estimation unit 251 (first biological waveform), and the biological waveform estimated by the second waveform estimation unit 252 (second biological waveform) on the display unit 22.

[0052] (4) Advantages As described above, the signal processing system 2 of Embodiment 1 comprises a signal acquisition unit 231, a waveform estimation unit 232, a waveform comparison unit 233, and a determination unit 234. The signal acquisition unit 231 acquires multiple detection signals (e.g., detection signals L1, L2) from the sensor system 10, each corresponding to a plurality of detection areas (e.g., detection areas G1, G2). The sensor system 10 is installed in space SP1 and detects people in a plurality of different detection areas. The waveform estimation unit 232 estimates bio-waveforms (e.g., a first bio-waveform and a second bio-waveform) related to a person's biological information from each of the plurality of detection signals. The waveform estimation unit 232 compares the plurality of bio-waveforms corresponding to each of the plurality of detection signals. The determination unit 234 determines the presence of a person in space SP1 based on the comparison result of the waveform comparison unit 233.

[0053] For example, a sensor system emits radio waves into each of its multiple detection areas and receives radio waves from each of these areas. It's possible that a radio wave reflected from one of the detection areas could be received from another detection area. In such a case, even if there is no person in one detection area, the system might mistakenly determine that there is a person in that area based on the radio waves received from other detection areas.

[0054] Therefore, as in Embodiment 1, the waveform estimation unit 232 can extract similarity by comparing multiple biological waveforms corresponding to multiple detection signals. This makes it possible to suppress a decrease in the accuracy of the determination even when determining the presence of a person in multiple detection areas using a person's biological information.

[0055] (5) Variant The following are examples of modifications. These modifications can be applied in appropriate combinations with Embodiment 1 described above.

[0056] (5.1) Variation 1 In Embodiment 1, the human detection system 1 is configured to determine whether there is one or two people in space SP1, but the system is not limited to this configuration.

[0057] The human detection system 1 may determine whether there is one, two, or three people in space SP1. The human detection system 1 of Modified Example 1 will be explained using Figures 9 and 10, focusing on the differences from Embodiment 1. Components similar to those in Embodiment 1 are denoted by the same reference numerals, and their descriptions are omitted as appropriate.

[0058] As shown in Figure 9, the modified example 1 of the human detection system 1 comprises a sensor system 10A and a signal processing device 20A as a signal processing system 2.

[0059] As shown in Figure 9, the sensor system 10A includes multiple (three in the illustrated example) radio wave sensors 11. When it is necessary to individually designate the multiple radio wave sensors 11, they are referred to as the first radio wave sensor 11a, the second radio wave sensor 11b, and the third radio wave sensor 11c.

[0060] Multiple radio wave sensors 11 detect biometric information (human respiration) of people (for example, people u1, u2, u3 shown in Figure 10) in multiple different detection areas G1, G2, and G3. For example, the sensor unit 101a of the first radio wave sensor 11a outputs radio waves into detection area G1 and receives radio waves reflected by objects in detection area G1. The sensor unit 101b of the second radio wave sensor 11b outputs radio waves into detection area G2 and receives radio waves reflected by objects in detection area G2. The sensor unit (not shown) of the third radio wave sensor 11c outputs radio waves into detection area G3 and receives radio waves reflected by objects in detection area G3. In the modified example 1, the detection areas G1, G2, and G3 are arranged in this order along a predetermined direction, as shown in Figure 10.

[0061] The configuration of the multiple radio wave sensors 11 is the same as that of the radio wave sensor 11 in Embodiment 1, so a detailed explanation is omitted here.

[0062] The signal generation unit 102 (see Figure 1) of the first radio wave sensor 11a generates a detection signal (Doppler signal) L1 based on an oscillation signal corresponding to the radio wave output by the sensor unit 101a and a reception signal corresponding to the radio wave received by the sensor unit 101 of the first radio wave sensor 11a. The signal generation unit 102 of the second radio wave sensor 11b generates a detection signal (Doppler signal) L2 based on an oscillation signal corresponding to the radio wave output by the sensor unit 101 of the second radio wave sensor 11b and a reception signal corresponding to the radio wave received by the sensor unit 101 of the second radio wave sensor 11b. The signal generation unit (not shown) of the third radio wave sensor 11c generates a detection signal (Doppler signal) L3 based on an oscillation signal corresponding to the radio wave output by the sensor unit (not shown) of the third radio wave sensor 11c and a reception signal corresponding to the radio wave received by the sensor unit of the third radio wave sensor 11c. Hereafter, detection signal L1 may be referred to as the first detection signal L1, detection signal L2 as the second detection signal L2, and detection signal L3 as the third detection signal L3.

[0063] The first radio wave sensor 11a outputs a first detection signal L1 to the signal processing device 20A. The second radio wave sensor 11b outputs a second detection signal L2 to the signal processing device 20A. The third radio wave sensor 11c outputs a third detection signal L3 to the signal processing device 20A.

[0064] As shown in Figure 9, the signal processing device 20A comprises a communication unit 21, a display unit 22, and a control unit 23A.

[0065] The signal processing device 20A includes, for example, a computer system having a processor and memory. The computer system functions as a control unit 23A when the processor executes a program stored in memory. The program executed by the processor is pre-recorded in the computer system's memory, but it may also be provided on a recording medium such as a memory card, or provided via a telecommunications line such as the Internet.

[0066] As shown in Figure 9, the control unit 23A includes a signal acquisition unit 231A, a waveform estimation unit 232A, a waveform comparison unit 233A, a determination unit 234A, and a display processing unit 235.

[0067] The signal acquisition unit 231A acquires multiple detection signals corresponding to each of the multiple detection areas G1, G2, and G3 from the sensor system 10, which is installed in space SP1 and detects people in multiple different detection areas G1, G2, and G3, via the communication unit 21. The signal acquisition unit 231 includes a first signal acquisition unit 241, a second signal acquisition unit 242, and a third signal acquisition unit 243. The first signal acquisition unit 241 acquires the first detection signal L1 output from the first radio wave sensor 11a. The second signal acquisition unit 242 acquires the second detection signal L2 output from the second radio wave sensor 11b. The third signal acquisition unit 243 acquires the third detection signal L3 output from the third radio wave sensor 11c.

[0068] The waveform estimation unit 232A estimates a bio-waveform related to human biological information from each of the multiple detection signals acquired by the signal acquisition unit 231A. In this case, the waveform estimation unit 232A estimates a human respiratory waveform from each of the multiple detection signals acquired by the signal acquisition unit 231. The waveform estimation unit 232A estimates multiple bio-waveforms corresponding to each of the multiple detection signals using a data estimation learning model constructed by a recurrent neural network.

[0069] The waveform estimation unit 232A includes a first waveform estimation unit 251, a second waveform estimation unit 252, and a third waveform estimation unit 253.

[0070] The first waveform estimation unit 251 estimates the biological waveform (respiratory waveform) based on the first detection signal L1 acquired by the first signal acquisition unit 241. The second waveform estimation unit 252 estimates the biological waveform (respiratory waveform) based on the second detection signal L2 acquired by the second signal acquisition unit 242. The third waveform estimation unit 253 estimates the biological waveform (respiratory waveform) based on the third detection signal L3 acquired by the third signal acquisition unit 243. In other words, the waveform estimation unit 232A estimates the biological waveform corresponding to the first detection signal L1 (first biological waveform), the biological waveform corresponding to the second detection signal L2 (second biological waveform), and the biological waveform corresponding to the third detection signal L3 (third biological waveform).

[0071] The first waveform estimation unit 251 estimates the human biological waveform from the first detection signal L1 using a learned model. The second waveform estimation unit 252 estimates the human biological waveform from the second detection signal L2 using the same learned model. The third waveform estimation unit 253 estimates the human biological waveform from the third detection signal L3 using the same learned model.

[0072] The waveform comparison unit 233A compares multiple biological waveforms corresponding to multiple detection signals. Specifically, the waveform comparison unit 233 compares a first biological waveform based on the first detection signal L1 with a second biological waveform based on the second detection signal L2. The waveform comparison unit 233 also compares a second biological waveform based on the second detection signal L2 with a third biological waveform based on the third detection signal L3.

[0073] The waveform comparison unit 233A calculates an evaluation index representing the similarity between the first and second biological waveforms as a result of comparing the first and second biological waveforms. For example, the evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship between multiple biological waveforms (e.g., the first and second biological waveforms) over a predetermined period. The waveform comparison unit 233 calculates multiple cross-correlation coefficients for multiple biological waveforms over a predetermined period using a cross-correlation function. The waveform comparison unit 233 extracts the calculated multiple cross-correlation coefficients as the first comparison result between the first and second biological waveforms.

[0074] Furthermore, the waveform comparison unit 233A calculates an evaluation index representing the similarity between the second and third biological waveforms as a comparison result between the second and third biological waveforms. For example, the evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship between multiple biological waveforms (e.g., the second and third biological waveforms) over a predetermined period. The waveform comparison unit 233 calculates multiple cross-correlation coefficients for multiple biological waveforms over a predetermined period using a cross-correlation function. The waveform comparison unit 233 extracts the calculated multiple cross-correlation coefficients as the first comparison result between the second and third biological waveforms.

[0075] The determination unit 234A determines the presence of a person in space SP1 based on the first and second comparison results of the waveform comparison unit 233A. For example, the determination unit 234A determines the presence of a person in space SP1 using the first and second comparison results of the waveform comparison unit 233A and Table 1.

[0076] [Table 1]

[0077] Specifically, the determination unit 234A determines that there are 3 people in space SP1 if the cross-correlation coefficient, which is the first comparison result, is smaller than a predetermined first threshold, and the cross-correlation coefficient, which is the second comparison result, is smaller than a predetermined second threshold. The determination unit 234A determines that there are 2 people in space SP1 if the cross-correlation coefficient, which is the first comparison result, is smaller than a predetermined first threshold, and the cross-correlation coefficient, which is the second comparison result, is greater than or equal to a predetermined second threshold. The determination unit 234A determines that there are 2 people in space SP1 if the cross-correlation coefficient, which is the first comparison result, is greater than or equal to a predetermined first threshold, and the cross-correlation coefficient, which is the second comparison result, is smaller than a predetermined second threshold. The determination unit 234A determines that there is 1 person in space SP1 if the cross-correlation coefficient, which is the first comparison result, is greater than or equal to a predetermined first threshold, and the cross-correlation coefficient, which is the second comparison result, is greater than or equal to a predetermined second threshold.

[0078] Here, a predetermined first threshold is a value based on the cross-correlation coefficient corresponding to the peak value in the frequency distribution of multiple cross-correlation coefficients calculated using the first person bio-waveform and the second person bio-waveform over a predetermined period. A predetermined second threshold is a value based on the cross-correlation coefficient corresponding to the peak value in the frequency distribution of multiple cross-correlation coefficients calculated using the second person bio-waveform and the third person bio-waveform over a predetermined period. The first person bio-waveform is a bio-waveform estimated by the first waveform estimation unit 251 when only the first person (e.g., person u1) is present in space SP1. The second person bio-waveform is a bio-waveform estimated by the second waveform estimation unit 252 when only the second person (e.g., person u2) is present in space SP1. The third person bio-waveform is a bio-waveform estimated by the third waveform estimation unit 253 when only the third person (e.g., person u3) is present in space SP1. For example, a predetermined first threshold is a value greater than the cross-correlation coefficient corresponding to the peak value in the frequency distribution of the multiple calculated cross-correlation coefficients. A predetermined second threshold is a value greater than the cross-correlation coefficient corresponding to the peak value in the frequency distribution of the multiple calculated cross-correlation coefficients.

[0079] The display processing unit 235 displays the determination result from the determination unit 234 on the display unit 22. Furthermore, the display processing unit 235 displays the biological waveform estimated by the first waveform estimation unit 251 (first biological waveform), the biological waveform estimated by the second waveform estimation unit 252 (second biological waveform), and the biological waveform estimated by the third waveform estimation unit 253 (third biological waveform) on the display unit 22.

[0080] (5.2) Variation 2 In Embodiment 1, the multiple radio wave sensors 11 provided by the human detection system 1 are configured to be, for example, Doppler sensors, FMCW type radio wave sensors, etc., but the system is not limited to this configuration.

[0081] The human detection system 1 may include a MIMO (Multiple Input Multiple Output) radio wave sensor. Below, the human detection system 1 of Modified Example 2 will be described with reference to Figure 11, focusing on the differences from Embodiment 1. Components similar to those in Embodiment 1 are denoted by the same reference numerals, and their descriptions are omitted as appropriate.

[0082] As shown in Figure 11, the modified human detection system 1 of the second example comprises a sensor system 10B and a signal processing device 20 as a signal processing system 2.

[0083] As shown in Figure 11, the sensor system 10B includes a MIMO-type radio wave sensor 12. The radio wave sensor 12 has multiple (e.g., two) transmitting antennas and multiple (e.g., two) receiving antennas. Furthermore, the radio wave sensor 12 includes a sensor unit 121, a beamforming unit 122, a signal generation unit 123, and a communication unit 124.

[0084] The sensor unit 121 detects biometric information (human respiration) of people (for example, people u1, u2, u3 shown in Figure 10) in multiple distinct detection areas G1, G2, and G3. For example, the sensor unit 121 uses multiple transmitting antennas to output radio waves into detection area G1 and detection area G2, respectively, and receives the radio waves reflected by objects in detection area G1 with multiple receiving antennas.

[0085] The beamforming unit 122 separates the received signal corresponding to the radio waves received by the sensor unit 121 into a first received signal corresponding to the radio waves reflected by an object in the detection area G1 and a second received signal corresponding to the radio waves reflected by an object in the detection area G2.

[0086] The signal generation unit 123 generates a first detection signal L1 based on a first oscillation signal corresponding to the radio waves output towards the detection area G1 and a first reception signal. The signal generation unit 123 generates a second detection signal L2 based on a second oscillation signal corresponding to the radio waves output towards the detection area G2 and a second reception signal. For example, the first detection signal L1 and the second detection signal L2 are Doppler signals that include in-phase and orthogonal components.

[0087] The communication unit 124 has a communication interface for communicating with the signal processing device 20. The communication unit 124 outputs (transmits) the first detection signal L1 and the second detection signal L2 generated by the signal generation unit 123 to the signal processing device 20.

[0088] In the modified version 2, the signal processing device 20 receives a first detection signal L1 and a second detection signal L2 from the sensor system 10B, and calculates a cross-correlation coefficient based on the received first detection signal L1 and second detection signal L2. The signal processing device 20 in the modified version 2 uses the calculated cross-correlation coefficient and a predetermined threshold TH1 to determine the presence of a person in space SP1.

[0089] (5.3) Modification 3 The waveform comparison unit 233 may calculate multiple cross-correlation coefficients for each of the multiple predetermined periods and extract the cross-correlation coefficient corresponding to the peak value in the frequency distribution of the multiple cross-correlation coefficients as the maximum cross-correlation coefficient. The waveform comparison unit 233 then uses the calculated value obtained from the multiple maximum cross-correlation coefficients extracted for each of the multiple predetermined periods as the comparison result. In this case, the determination unit 234 determines the presence of a person in space SP1 based on the calculated value.

[0090] Here, the calculated value is one of the mean, median, maximum, or minimum values ​​of multiple maximum cross-correlation coefficients.

[0091] (5.4) Modification 4 The predetermined threshold may be a value based on the cross-correlation coefficient corresponding to the peak value in the frequency distribution obtained by combining multiple first-criterion cross-correlation coefficients and multiple second-criterion cross-correlation coefficients. The multiple first-criterion cross-correlation coefficients are multiple criterion cross-correlation coefficients over a predetermined period in a state where only a first person (e.g., person u1) exists in space SP1. The multiple second-criterion cross-correlation coefficients are multiple criterion cross-correlation coefficients over a predetermined period in a state where only a second person (e.g., person u2) exists in space SP1.

[0092] Here, multiple first criterion cross-correlation coefficients are obtained from the first waveform, which is the first biological waveform estimated by the first waveform estimation unit 251 for a predetermined period, and the second waveform, which is the second biological waveform estimated by the second waveform estimation unit 252 for a predetermined period, when only the first person is present in space SP1. Furthermore, multiple second criterion cross-correlation coefficients are obtained from the third waveform, which is the first biological waveform estimated by the first waveform estimation unit 251 for a predetermined period, and the fourth waveform, which is the second biological waveform estimated by the second waveform estimation unit 252 for a predetermined period, when only the second person is present in space SP1. In other words, the waveform comparison unit 233 calculates multiple first criterion cross-correlation coefficients for a predetermined period from the first waveform and the second waveform. The waveform comparison unit 233 calculates multiple second criterion cross-correlation coefficients for a predetermined period from the third waveform and the fourth waveform.

[0093] (5.5) Variation 5 The display processing unit 235 may display the determination result from the determination unit 234, the bio-waveform estimated by the first waveform estimation unit 251 (first bio-waveform), and the bio-waveform estimated by the second waveform estimation unit 252 (second bio-waveform) on an external terminal such as an information terminal having a display unit. Here, the information terminal is a smartphone, tablet terminal, etc.

[0094] (5.6) Variation 6 The determination unit 234 may determine the presence of a person using the following equation 2.

[0095]

number

[0096] Here, the parameter TH of Equation 2 prob is a parameter for setting the boundary of amplitude suppression. Also, the parameter g is a parameter for setting the intensity of suppression. Further, XC 12 is a variable representing the cross-correlation coefficient.

[0097] By using Equation 2 above, the line segment P1 shown in FIG. 12 is obtained. Here, the horizontal axis in FIG. 12 represents XC 12 and the vertical axis represents the value (J prob ) obtained from Equation 2. The value (J prob ) represents the probability that two people are in the space SP1.

[0098] The determination unit 234 determines the presence of people in the space SP1 based on the value (J 12 ) obtained by substituting the cross-correlation coefficient, which is the comparison result of the waveform comparison unit 233, into the variable XC prob . Specifically, if the value (J prob ) is 0.5 or more, the determination unit 234 determines that two people are in the space SP1. If the value (J prob ) is less than 0.5, the determination unit 234 determines that one person is in the space SP1.

[0099] (Embodiment 2) In Embodiment 2, it is different from Embodiment 1 in that the determination unit 234 adjusts the biological waveform according to the determination result. Hereinafter, the description will focus on the different points. For the same components as in Embodiment 1, the same reference numerals are given, and the description thereof will be omitted as appropriate.

[0100] (1) Configuration The human detection system 1 according to Embodiment 2 includes, as shown in FIG. 13, a sensor system 10 and a signal processing device 20C as a signal processing system 2.

[0101] The signal processing device 20C includes, as shown in FIG. 13, a communication unit 21, a display unit 22, and a control unit 23C.

[0102] The signal processing device 20C includes, for example, a computer system having a processor and memory. The computer system functions as a control unit 23C when the processor executes a program stored in memory. The program executed by the processor is pre-recorded in the computer system's memory, but it may also be provided on a recording medium such as a memory card, or provided via a telecommunications line such as the Internet.

[0103] As shown in Figure 13, the control unit 23C includes a signal acquisition unit 231, a waveform estimation unit 232, a waveform comparison unit 233, a determination unit 234, a display processing unit 235, and a waveform adjustment unit 236.

[0104] The waveform adjustment unit 236 adjusts the biological waveform according to the determination result of the determination unit 234.

[0105] If the determination unit 234 determines that there is one person in space SP1, the waveform adjustment unit 236 calculates the first average amplitude of the first biological waveform and the second average amplitude of the second biological waveform. The waveform adjustment unit 236 adjusts the amplitude of the biological waveform with a small average amplitude among the first and second biological waveforms by multiplying it by a suppression coefficient, but does not adjust the amplitude of the biological waveform with a large average amplitude among the first and second biological waveforms.

[0106] Specifically, the waveform adjustment unit 236 adjusts the amplitude of the bio-waveform with the smaller average amplitude among the first and second bio-waveforms, using a suppression coefficient corresponding to the cross-correlation coefficient value, which is the comparison result of the waveform comparison unit 233. For example, the waveform adjustment unit 236 uses the line segment P11 (see Figure 14), which represents the relationship between the cross-correlation coefficient and the suppression coefficient, to determine the suppression coefficient corresponding to the cross-correlation coefficient value, which is the comparison result of the waveform comparison unit 233. The waveform adjustment unit 236 multiplies the amplitude of the bio-waveform with the smaller average amplitude among the first and second bio-waveforms by the determined suppression coefficient.

[0107] If the determination unit 234 determines that there are two people in space SP1, the waveform adjustment unit 236 does not adjust the amplitude of either the first biological waveform or the second biological waveform.

[0108] Here, line segment P11 is obtained using equation 3 shown below. The parameter TH in equation 3 is 12 This parameter sets the boundary for amplitude suppression. The parameter g sets the intensity of the suppression. Furthermore, XC 12 This is a variable that represents the cross-correlation coefficient.

[0109]

number

[0110] By using Mathematics 3, if we assume that there is one person in space SP1, the amplitude of one of the first and second biological waveforms (the waveform with the smaller average amplitude) can be adjusted (suppressed) by a suppression coefficient corresponding to the cross-correlation coefficient.

[0111] If the determination unit 234 determines that there is one person in space SP1, the display processing unit 235 displays on the display unit 22 the first bio-waveform and the second bio-waveform whose amplitude has been adjusted, as well as the first bio-waveform and the second bio-waveform whose amplitude has not been adjusted and whose average amplitude is large. If the determination unit 234 determines that there are two people in space SP1, the display processing unit 235 displays on the display unit 22 the first bio-waveform and the second bio-waveform whose amplitude has not been adjusted.

[0112] For example, when the waveform estimation unit 232 estimates the first biological waveform W11 and the second biological waveform W12 shown in Figure 15A, the cross-correlation coefficient, which is the comparison result from the waveform comparison unit 233, becomes smaller than a predetermined value. As a result, the determination unit 234 determines that there are two people in space SP1. In this case, the waveform adjustment unit 236 outputs the first biological waveform W21 and the second biological waveform W22 without adjusting the waveforms of both the first biological waveform W11 and the second biological waveform W12. That is, the first biological waveform W11 and the first biological waveform W21 shown in Figure 15A are the same waveform, and the second biological waveform W12 and the second biological waveform W22 shown in Figure 15A are the same waveform. The display processing unit 235 displays the first biological waveform W21 and the second biological waveform W22, i.e., the first biological waveform W11 and the second biological waveform W12, output from the waveform adjustment unit 236 on the display unit 22.

[0113] Furthermore, when the waveform estimation unit 232 estimates the first biological waveform W31 and the second biological waveform W32 shown in Figure 15B, the cross-correlation coefficient, which is the comparison result from the waveform comparison unit 233, will be greater than or equal to a predetermined value. As a result, the determination unit 234 determines that there is one person in space SP1. In this case, the waveform adjustment unit 236 calculates the first average amplitude of the first biological waveform W31 and the second average amplitude of the second biological waveform W32. The waveform adjustment unit 236 uses a suppression coefficient corresponding to the value of the cross-correlation coefficient, which is the comparison result from the waveform comparison unit 233, to adjust the amplitude of the biological waveform with the smaller average amplitude among the first and second biological waveforms. For example, if the second average amplitude is less than or equal to the first average amplitude, the waveform adjustment unit 236 adjusts the amplitude of the second biological waveform W32 using the suppression coefficient. The waveform adjustment unit 236 does not adjust the amplitude of the first biological waveform W31. The waveform adjustment unit 236 outputs a biological waveform with adjusted amplitude (here, the second biological waveform W42) and a biological waveform with unadjusted amplitude (here, the first biological waveform W41). That is, the first biological waveform W31 and the first biological waveform W41 shown in Figure 15B are the same waveform. The display processing unit 235 displays the first biological waveform W41 and the second biological waveform W42 output from the waveform adjustment unit 236 on the display unit 22.

[0114] (2) Operation The operation of the signal processing device 20C will be explained using Figure 16.

[0115] The signal acquisition unit 231 acquires multiple detection signals from the sensor system 10 corresponding to multiple detection areas G1 and G2 (see Figure 2) (step S101). Specifically, the first signal acquisition unit 241 of the signal acquisition unit 231 acquires the first detection signal L1 output from the first radio wave sensor 11a. The second signal acquisition unit 242 of the signal acquisition unit 231 acquires the second detection signal L2 output from the second radio wave sensor 11b.

[0116] The waveform estimation unit 232 performs waveform estimation processing (step S102). Specifically, the first waveform estimation unit 251 of the waveform estimation unit 232 estimates a first biological waveform based on the first detection signal L1 acquired by the first signal acquisition unit 241. The second waveform estimation unit 252 of the waveform estimation unit 232 estimates a second biological waveform based on the second detection signal L2 acquired by the second signal acquisition unit 242.

[0117] The waveform comparison unit 233 performs waveform comparison processing (step S103). Specifically, the waveform comparison unit 233 compares the first biological waveform with the second biological waveform. More specifically, the waveform comparison unit 233 uses a cross-correlation function to calculate multiple cross-correlation coefficients for multiple biological waveforms over a predetermined period.

[0118] The determination unit 234 uses the comparison result (overall correlation coefficient) from the waveform comparison unit 233 to determine the presence of a person in space SP1 (step S104). Here, the determination unit 234 uses the comparison result (overall correlation coefficient) from the waveform comparison unit 233 to determine whether there are two people or one person in space SP1. The determination unit 234 determines the presence of a person in space SP1 based on the cross-correlation coefficient extracted as a comparison result by the waveform comparison unit 233. Specifically, if the cross-correlation coefficient, which is the comparison result, is above a predetermined threshold, the determination unit 234 determines that there is one person in space SP1. If the cross-correlation coefficient, which is the comparison result, is below a predetermined threshold, the determination unit 234 determines that there are two people in space SP1.

[0119] If the determination unit 234 determines that there are two people in space SP1 ("Yes" in step S104), the display processing unit 235 performs the first display processing (step S105). The display processing unit 235 displays the first and second biological waveforms, which have not undergone amplitude adjustment, on the display unit 22.

[0120] If the determination unit 234 determines that there are no two people in space SP1, that is, that there is one person in space SP1 ("No" in step S104), the waveform adjustment unit 236 calculates the first average amplitude of the first biological waveform and the second average amplitude of the second biological waveform (step S106).

[0121] The waveform adjustment unit 236 determines the suppression coefficient (step S107). For example, the waveform adjustment unit 236 uses the line segment P11 (see Figure 14), which represents the relationship between the cross-correlation coefficient and the suppression coefficient, i.e., equation 3 above, to determine the suppression coefficient corresponding to the value of the cross-correlation coefficient, which is the comparison result of the waveform comparison unit 233.

[0122] The waveform adjustment unit 236 determines whether the first average amplitude is greater than or equal to the second average amplitude based on the calculation result in step S106 (step S108).

[0123] If it is determined that the first average amplitude is greater than or equal to the second average amplitude (Yes in step S108), the waveform adjustment unit 236 adjusts the amplitude of the second biological waveform, which has a smaller average amplitude than the first biological waveform (step S109). Specifically, the waveform adjustment unit 236 adjusts the amplitude of the second biological waveform by multiplying its amplitude by a suppression coefficient.

[0124] If it is determined that the first average amplitude is not greater than or equal to the second average amplitude (No in step S108), the waveform adjustment unit 236 adjusts the amplitude of the first biological waveform, which has a smaller average amplitude than the first biological waveform (step S110). Specifically, the waveform adjustment unit 236 adjusts the amplitude of the second biological waveform by multiplying the amplitude of the first biological waveform by the suppression coefficient determined in step S107.

[0125] The display processing unit 235 performs a second display process (step S111). The display processing unit 235 displays on the display unit 22 the biological waveforms whose amplitudes have been adjusted and the biological waveforms whose amplitudes have not been adjusted from the first biological waveform and the second biological waveform. Specifically, if step S109 is executed, the display processing unit 235 displays on the display unit 22 the second biological waveform whose amplitudes have been adjusted and the first biological waveform whose amplitudes have not been adjusted. Also, if step S110 is executed, the display processing unit 235 displays on the display unit 22 the first biological waveform whose amplitudes have been adjusted and the second biological waveform whose amplitudes have not been adjusted.

[0126] (3) Advantages In Embodiment 2, as in Embodiment 1, it is possible to suppress a decrease in the accuracy of the determination even when determining the presence of a person in multiple detection areas using a person's biometric information.

[0127] Furthermore, the determination unit 234 can also make determinations as shown in Table 2 below by comparing the average amplitudes of multiple biological waveforms.

[0128] [Table 2]

[0129] For example, the determination unit 234 determines that there are two people in space SP1 if the comparison result of the waveform comparison unit 233 is smaller than a predetermined threshold and the difference between the first average amplitude and the second average amplitude is smaller than a predetermined value.

[0130] The determination unit 234 determines that there is one person in space SP1 if the comparison result of the waveform comparison unit 233 is greater than or equal to a predetermined threshold and the difference between the first average amplitude and the second average amplitude is greater than or equal to a predetermined value. In this case, if the first average amplitude is greater than the second average amplitude, the determination unit 234 determines that there is one person in the detection area G1. If the second average amplitude is greater than the first average amplitude, the determination unit 234 determines that there is one person in the detection area G2.

[0131] Furthermore, if the comparison result of the waveform comparison unit 233 is greater than or equal to a predetermined threshold, and the difference between the first average amplitude and the second average amplitude is less than a predetermined value, the determination unit determines that one person exists in space SP1, spanning both detection area G1 and detection area G2.

[0132] Thus, the signal processing system 2 of Embodiment 2 can accurately determine the presence of a person even when that person is located across detection area G1 and detection area G2.

[0133] (4) Variations The following are examples of modifications. These modifications can be applied in appropriate combination with Embodiment 2 described above.

[0134] (4.1) Variation 1 The value obtained using the number 2 in modified example 6 of Embodiment 1 (J prob ) may be used as the suppression coefficient. The operation of the signal processing system 2 (signal processing device 20C) in modified example 1 of Embodiment 2 will be explained with reference to Figure 17.

[0135] The signal acquisition unit 231 acquires multiple detection signals from the sensor system 10 corresponding to multiple detection areas G1 and G2 (see Figure 2) (step S201). Specifically, the first signal acquisition unit 241 of the signal acquisition unit 231 acquires the first detection signal L1 output from the first radio wave sensor 11a. The second signal acquisition unit 242 of the signal acquisition unit 231 acquires the second detection signal L2 output from the second radio wave sensor 11b.

[0136] The waveform estimation unit 232 performs waveform estimation processing (step S202). Specifically, the first waveform estimation unit 251 of the waveform estimation unit 232 estimates a first biological waveform based on the first detection signal L1 acquired by the first signal acquisition unit 241. The second waveform estimation unit 252 of the waveform estimation unit 232 estimates a second biological waveform based on the second detection signal L2 acquired by the second signal acquisition unit 242.

[0137] The waveform comparison unit 233 performs waveform comparison processing (step S203). Specifically, the waveform comparison unit 233 compares the first biological waveform with the second biological waveform. More specifically, the waveform comparison unit 233 uses a cross-correlation function to calculate multiple cross-correlation coefficients for multiple biological waveforms over a predetermined period.

[0138] The determination unit 234 uses the comparison result (overall correlation coefficient) from the waveform comparison unit 233 to determine the presence of a person in space SP1 (step S204). Here, the determination unit 234 uses the comparison result (overall correlation coefficient) from the waveform comparison unit 233 and equation 2 to determine the suppression coefficient (J prob The determination unit 234 calculates the calculated suppression coefficient (J). prob The presence of a person in space SP1 is determined based on the suppression coefficient (J). For example, the determination unit 234 determines the presence of a person based on the suppression coefficient (J). prob If the suppression coefficient (J) is less than 0.5, it is determined that there is one person in space SP1. The determination unit 234 determines the suppression coefficient (J) prob If the value is 0.5 or greater, it is determined that there are two people in space SP1.

[0139] If the determination unit 234 determines that there are two people in space SP1 ("Yes" in step S204), the waveform adjustment unit 236 performs the first adjustment process (step S205). Specifically, the waveform adjustment unit 236 applies a suppression coefficient (J) to both the first biological waveform and the second biological waveform. prob The amplitudes of both the first and second bio-waveforms are adjusted by multiplying by ).

[0140] The display processing unit 235 performs the first display processing (step S206). The display processing unit 235 displays the first and second biological waveforms, whose amplitudes have been adjusted in the first adjustment processing, on the display unit 22.

[0141] If the determination unit 234 determines that there are no two people in space SP1, that is, that there is one person in space SP1 ("No" in step S204), the waveform adjustment unit 236 calculates the first average amplitude of the first biological waveform and the second average amplitude of the second biological waveform (step S207).

[0142] The waveform adjustment unit 236 determines whether the first average amplitude is greater than or equal to the second average amplitude based on the calculation result in step S207 (step S208).

[0143] If it is determined that the first average amplitude is greater than or equal to the second average amplitude (Yes in step S208), the waveform adjustment unit 236 performs the second adjustment process (step S209). Specifically, the waveform adjustment unit 236 applies a suppression coefficient (J) to the amplitude of the second biological waveform, which has a smaller average amplitude than the first biological waveform. prob The amplitude of the second biological waveform is adjusted by multiplying it by ). The waveform adjustment unit 236 sets the amplitude of the first biological waveform, which has a larger average amplitude than the first biological waveform, to a value "1 - suppression coefficient (J)". prob The amplitude of the first bio-waveform is adjusted by multiplying by ").

[0144] If it is determined that the first average amplitude is not greater than or equal to the second average amplitude (No in step S208), the waveform adjustment unit 236 performs a third adjustment process (step S210). Specifically, the waveform adjustment unit 236 applies a suppression coefficient (J) to the amplitude of the first biological waveform, which has a smaller average amplitude than the second biological waveform. prob The amplitude of the first biological waveform is adjusted by multiplying it by ). The waveform adjustment unit 236 sets the amplitude of the second biological waveform, which has a larger average amplitude than the first biological waveform, to a value "1 - suppression coefficient (J)". prob Multiply by ")" to adjust the amplitude of the second bio-waveform.

[0145] The display processing unit 235 performs a second display processing (step S211). Specifically, if step S209 is executed, the display processing unit 235 calculates the suppression coefficient (J prob The second bio-waveform is multiplied by ) and the value "1 - suppression coefficient (J) prob The first biological waveform multiplied by the suppression coefficient (J) is displayed on the display unit 22. If step S210 is executed, the display processing unit 235 displays the suppression coefficient (J) prob The first biological waveform is multiplied by ) and the value "1 - suppression coefficient (J) prob The second biological waveform, multiplied by ")", is displayed on the display unit 22.

[0146] (4.2) Modification 2 The waveform adjustment unit 236 may identify a biological waveform from among the first and second biological waveforms whose amplitude should be adjusted, based on the amplitudes of the multiple Doppler signals L1 and L2 output from the sensor system 10.

[0147] (4.3) Modification 3 The modifications 1 to 5 of Embodiment 1 may be applied to Embodiment 2 as appropriate.

[0148] (Other variations) The above embodiments are merely one of many embodiments of this disclosure. The above embodiments can be modified in various ways depending on the design, etc., as long as the objectives of this disclosure are achieved. Furthermore, functions similar to those of the signal processing system 2 may be embodied in a signal processing method, a computer program, or a non-temporary recording medium on which a program is recorded. A signal processing method of the signal processing system 2 according to one embodiment includes a signal acquisition step, a waveform estimation step, a waveform comparison step, and a determination step. The signal acquisition step acquires multiple detection signals (e.g., detection signals L1, L2) corresponding to multiple detection areas G1, G2, respectively, from sensor systems 10, 10A, 10B, which are provided in space SP1 and detect people in multiple different detection areas G1, G2. The waveform estimation step estimates a bio-waveform relating to a person's biological information from each of the multiple detection signals. The waveform comparison step compares multiple bio-waveforms corresponding to each of the multiple detection signals. The determination step determines the presence of a person in space SP1 based on the comparison result in the waveform comparison step. A program according to one embodiment is a program for causing a computer system to function as the above-described signal processing system 2 or as a signal processing method of the signal processing system 2.

[0149] The entity executing the signal processing system 2 or the signal processing method of the signal processing system 2 in this disclosure includes a computer system. The computer system has a processor and memory as hardware. The processor executes a program recorded in the memory of the computer system, thereby realizing the function of the entity executing the signal processing system 2 or the signal processing method of the signal processing system 2 in this disclosure. The program may be pre-recorded in the memory of the computer system, or it may be provided via a telecommunications line. Alternatively, the program may be recorded and provided on a non-temporary recording medium such as a memory card, optical disk, or hard disk drive that is readable by the computer system. The processor of the computer system is composed of one or more electronic circuits including semiconductor integrated circuits (ICs) or large-scale integrated circuits (LSIs). The integrated circuits such as ICs or LSIs referred to here are named differently depending on the degree of integration, and include integrated circuits called system LSIs, VLSIs (Very Large Scale Integration), or ULSIs (Ultra Large Scale Integration). Furthermore, FPGAs (Field-Programmable Gate Arrays) that are programmed after the LSI is manufactured, or logic devices capable of reconfiguring the junction relationships or circuit compartments within the LSI, can also be used as processors. Multiple electronic circuits may be integrated onto a single chip or distributed across multiple chips. Multiple chips may be integrated into a single device or distributed across multiple devices.

[0150] Furthermore, it is not essential for the signal processing system 2 to have multiple functions integrated into a single enclosure; the components of the signal processing system 2 may be distributed across multiple enclosures. Moreover, at least some of the functions of the signal processing system 2, for example, some of the functions of the signal processing system 2, may be implemented by the cloud (cloud computing), etc.

[0151] (summary) As described above, the signal processing system (2) of the first embodiment comprises a signal acquisition unit (231), a waveform estimation unit (232; 232A), a waveform comparison unit (233; 233A), and a determination unit (234; 234A). The signal acquisition unit (231) acquires multiple detection signals (e.g., detection signals L1, L2) corresponding to multiple detection areas (G1, G2) from a sensor system (10; 10A; 10B). The sensor system (10; 10A; 10B) is installed in space (SP1) and detects people in multiple different detection areas (G1, G2). The waveform estimation unit (232; 232A) estimates bio-waveforms (e.g., a first bio-waveform and a second bio-waveform) relating to a person's biological information from each of the multiple detection signals. The waveform estimation unit (232; 232A) compares multiple bio-waveforms corresponding to each of the multiple detection signals. The determination unit (234;234A) determines the presence of a person in space (SP1) based on the comparison results of the waveform comparison unit (233;233A).

[0152] According to this embodiment, even when using a person's biometric information to determine the presence of a person in multiple detection areas (G1, G2), it is possible to suppress a decrease in the accuracy of the determination.

[0153] In the second embodiment of the signal processing system (2), in the first embodiment, the waveform estimation unit (232; 232A) estimates multiple biological waveforms corresponding to multiple detection signals using a learning model for data estimation constructed by a recurrent neural network.

[0154] According to this embodiment, the accuracy of estimating biological waveforms can be improved by using a learning model constructed with a recurrent neural network.

[0155] In the signal processing system (2) of the third embodiment, in the first or second embodiment, a waveform comparison unit (233; 233A) calculates an evaluation index representing the similarity of multiple biological waveforms as a comparison result. A determination unit (234; 234A) determines the presence of a person in space (SP1) based on the evaluation index.

[0156] According to this embodiment, since the presence of a person is determined using an evaluation index that represents similarity, it is possible to suppress a decrease in the accuracy of the determination.

[0157] In the signal processing system (2) of the fourth embodiment, in the third embodiment, the evaluation index is a cross-correlation coefficient representing the cross-correlation relationship of multiple biological waveforms over a predetermined period. The waveform comparison unit (233;233A) uses a cross-correlation function to calculate multiple cross-correlation coefficients for multiple biological waveforms over a predetermined period as comparison results. The determination unit (234;234A) determines the presence of a person in space (SP1) based on the cross-correlation coefficients calculated by the waveform comparison unit (233;233A).

[0158] According to this embodiment, even when using a person's biometric information to determine the presence of a person in multiple detection areas (G1, G2), it is possible to suppress a decrease in the accuracy of the determination.

[0159] In the fifth aspect of the signal processing system (2), in the fourth aspect, the waveform comparison unit (233; 233A) calculates multiple cross-correlation coefficients for each of a plurality of predetermined periods, extracts the cross-correlation coefficient corresponding to the peak value in the frequency distribution of the plurality of cross-correlation coefficients as the maximum cross-correlation coefficient, and uses the calculated value obtained from the extracted maximum cross-correlation coefficients as the comparison result. The determination unit (234; 234A) determines the presence of a person in space (SP1) based on the calculated value.

[0160] According to this embodiment, even when using a person's biometric information to determine the presence of a person in multiple detection areas (G1, G2), it is possible to suppress a decrease in the accuracy of the determination.

[0161] In the signal processing system of the sixth embodiment (2), in the fifth embodiment, the calculated value is one of the mean, median, maximum, and minimum values ​​of a plurality of maximum cross-correlation coefficients.

[0162] According to this embodiment, the calculated value can be any of the mean, median, maximum, and minimum values ​​of multiple maximum cross-correlation coefficients.

[0163] In the signal processing system (2) of the seventh embodiment, in any of the fourth to sixth embodiments, the plurality of detection signals include a first detection signal (L1) and a second detection signal (L2). The waveform estimation unit (232; 232A) estimates a first biological waveform corresponding to the first detection signal (L1) and a second biological waveform corresponding to the second detection signal (L2). The waveform comparison unit (233; 233A) calculates a plurality of cross-correlation coefficients for the first biological waveform and the second biological waveform over a predetermined period as comparison results. The determination unit (234; 234A) determines whether there is one person or two people in the space (SP1) based on the comparison results and a predetermined threshold.

[0164] According to this embodiment, even when using a person's biometric information to determine the presence of a person in multiple detection areas (G1, G2), it is possible to suppress a decrease in the accuracy of the determination.

[0165] In the signal processing system (2) of the eighth embodiment, in the seventh embodiment, a predetermined threshold is a value based on a reference cross-correlation coefficient corresponding to the peak value in the frequency distribution of a plurality of reference cross-correlation coefficients among a plurality of reference cross-correlation coefficients calculated using the first human bio-waveform and the second human bio-waveform over a predetermined period. The first human bio-waveform is a waveform estimated by the waveform estimation unit (232;232A) when only the first person (person u1) is present in space (SP1). The second human bio-waveform is a waveform estimated by the waveform estimation unit (232;232A) when only the second person (person u2) is present in space (SP1).

[0166] According to this configuration, a predetermined threshold can be set according to the person being detected. Therefore, the accuracy of the detection can be improved.

[0167] In the signal processing system (2) of the ninth embodiment, in the seventh embodiment, the waveform estimation unit (232; 232A) includes a first waveform estimation unit (251) for estimating a first biological waveform and a second waveform estimation unit (252) for estimating a second biological waveform. A predetermined threshold is a value based on a reference cross-correlation coefficient corresponding to the peak value in a frequency distribution obtained by combining a plurality of first reference cross-correlation coefficients and a plurality of second reference cross-correlation coefficients. The plurality of first reference cross-correlation coefficients are a plurality of reference cross-correlation coefficients over a predetermined period in a state in which only a first person (person u1) exists in space (SP1). The plurality of second reference cross-correlation coefficients are a plurality of reference cross-correlation coefficients over a predetermined period in a state in which only a second person (person u2) exists in space (SP1). Multiple first-reference cross-correlation coefficients are obtained from the first waveform, which is the first biological waveform estimated by the first waveform estimation unit (251) for a predetermined period, and the second waveform, which is the second biological waveform estimated by the second waveform estimation unit (252) for a predetermined period, when only the first person (person u1) is present in space (SP1). Multiple second-reference cross-correlation coefficients are obtained from the third waveform, which is the first biological waveform estimated by the first waveform estimation unit (251) for a predetermined period, and the fourth waveform, which is the second biological waveform estimated by the second waveform estimation unit (252) for a predetermined period, when only the second person (person u2) is present in space (SP1).

[0168] According to this configuration, a predetermined threshold can be set according to the person being detected. Therefore, the accuracy of the detection can be improved.

[0169] The signal processing system (2) of the tenth embodiment further comprises a waveform adjustment unit (236) and a display processing unit (235) in the seventh embodiment. The waveform adjustment unit (236) calculates the first average amplitude of the first biological waveform and the second average amplitude of the second biological waveform when the determination unit (234; 234A) determines that there is one person in the space (SP1). The waveform adjustment unit (236) adjusts the amplitude of the biological waveform with a small average amplitude among the first and second biological waveforms by multiplying it by a suppression coefficient to suppress the amplitude. The waveform adjustment unit (236) does not adjust the amplitude of the biological waveform with a large average amplitude among the first and second biological waveforms. If the determination unit (234; 234A) determines that there are two people in the space (SP1), no amplitude adjustment is performed on either the first or second biological waveform. If the determination unit (234; 234A) determines that there is one person in the space (SP1), the display processing unit (235) displays on the display unit (22) the first and second biological waveforms whose amplitudes have been adjusted, and the biological waveform with a large average amplitude that has not been adjusted, both of the first and second biological waveforms. If the determination unit (234; 234A) determines that there are two people in the space (SP1), the display processing unit (235) displays on the display unit (22) the first and second biological waveforms whose amplitudes have not been adjusted.

[0170] According to this embodiment, it is possible to clearly distinguish between biological waveforms that have been incorrectly detected and those that have been detected correctly. Therefore, when visually inspecting each biological waveform, the possibility of misidentification can be reduced.

[0171] In the 11th embodiment of the signal processing system (2), in any of the first to ten embodiments, the biological waveform is a human respiratory waveform.

[0172] According to this embodiment, the presence of a person in space (SP1) can be determined based on their breathing.

[0173] The twelfth embodiment of the human detection system (1) comprises a signal processing system (2) of any of the first to eleventh embodiments and a sensor system (10; 10A; 10B) for detecting a person in a space (SP1).

[0174] According to this embodiment, even when using a person's biometric information to determine the presence of a person in multiple detection areas (G1, G2), it is possible to suppress a decrease in the accuracy of the determination.

[0175] A signal processing method according to the 13th embodiment includes a signal acquisition step, a waveform estimation step, a waveform comparison step, and a determination step. The signal acquisition step acquires multiple detection signals (e.g., detection signals L1, L2) corresponding to multiple detection areas (G1, G2) from a sensor system (10; 10A; 10B) installed in space (SP1) that detects people in multiple different detection areas (G1, G2). The waveform estimation step estimates a bio-waveform relating to a person's biological information from each of the multiple detection signals. The waveform comparison step compares multiple bio-waveforms corresponding to each of the multiple detection signals. The determination step determines the presence of a person in space (SP1) based on the comparison results in the waveform comparison step.

[0176] According to this embodiment, even when using a person's biometric information to determine the presence of a person in multiple detection areas (G1, G2), it is possible to suppress a decrease in the accuracy of the determination.

[0177] The program of the 14th embodiment is a program that causes a computer system to execute the signal processing method of the 13th embodiment.

[0178] According to this embodiment, even when using a person's biometric information to determine the presence of a person in multiple detection areas (G1, G2), it is possible to suppress a decrease in the accuracy of the determination. [Explanation of Symbols]

[0179] 1-person detection system 2. Signal Processing System 10, 10A, 10B Sensor System 20, 20A, 20C Signal Processing Device 22 Display section 232,232A Waveform estimation section 233,233A waveform comparison section 234,234A Judgment section 235 Display Processing Unit 236 Waveform adjustment section 251 1st waveform estimation section 252 Second waveform estimation section 253 Third waveform estimation section G1, G2 detection area G3 detection area L1 detection signal (first detection signal, Doppler signal) L2 detection signal (second detection signal, Doppler signal) L3 detection signal (third detection signal, Doppler signal) SP1 Space u1,u2,u3 people W1, W11, W21, W31, W41 Bio-waveform (First Bio-waveform) W2, W12, W22, W32, W42 Bio-waveform (Second Bio-waveform)

Claims

1. A signal acquisition unit that acquires multiple detection signals corresponding to each of the multiple detection areas from a sensor system installed in space that detects people in multiple different detection areas, A waveform estimation unit estimates a bio-waveform related to human biological information from each of the aforementioned plurality of detection signals, A waveform comparison unit that compares a plurality of biological waveforms corresponding to each of the plurality of detection signals, The system includes a determination unit that determines the number of people in the space based on the comparison results of the waveform comparison unit, The plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. The waveform comparison unit calculates an evaluation index representing the similarity of the multiple biological waveforms as a result of the comparison, The determination unit determines the number of people in the space based on the evaluation index, The evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of the multiple biological waveforms over a predetermined period. The waveform comparison unit uses a cross-correlation function to calculate a plurality of cross-correlation coefficients for the plurality of biological waveforms over the predetermined period as the comparison result. The determination unit determines the number of people in the space based on the cross-correlation coefficient calculated by the waveform comparison unit. The waveform comparison unit calculates multiple cross-correlation coefficients for each of the multiple predetermined periods, extracts the cross-correlation coefficient corresponding to the peak value in the frequency distribution of the multiple cross-correlation coefficients as the maximum cross-correlation coefficient, and uses the calculated value obtained from the extracted maximum cross-correlation coefficients as the comparison result. The determination unit determines the number of people in the space based on the calculated value. Signal processing system.

2. The calculated value is one of the mean, median, maximum, and minimum values ​​of the plurality of maximum cross-correlation coefficients. The signal processing system according to claim 1.

3. The plurality of detection signals include a first detection signal and a second detection signal. The waveform estimation unit estimates a first biological waveform corresponding to the first detection signal and a second biological waveform corresponding to the second detection signal. The waveform comparison unit calculates a plurality of cross-correlation coefficients for the first biological waveform and the second biological waveform over the predetermined period as the comparison result. The determination unit, Based on the comparison results and a predetermined threshold, it is determined whether one person or two people are present in the space. The signal processing system according to claim 1.

4. The plurality of detection signals include a first detection signal and a second detection signal. The waveform estimation unit estimates a first biological waveform corresponding to the first detection signal and a second biological waveform corresponding to the second detection signal. The waveform comparison unit calculates a plurality of cross-correlation coefficients for the first biological waveform and the second biological waveform over the predetermined period as the comparison result. The determination unit, Based on the comparison results and a predetermined threshold, it is determined whether one person or two people are present in the space. The signal processing system according to claim 2.

5. A signal acquisition unit that acquires multiple detection signals corresponding to each of the multiple detection areas from a sensor system installed in space that detects people in multiple different detection areas, A waveform estimation unit estimates a bio-waveform related to human biological information from each of the aforementioned plurality of detection signals, A waveform comparison unit that compares a plurality of biological waveforms corresponding to each of the plurality of detection signals, The system includes a determination unit that determines the number of people in the space based on the comparison results of the waveform comparison unit, The plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. The waveform comparison unit calculates an evaluation index representing the similarity of the multiple biological waveforms as a result of the comparison, The determination unit determines the number of people in the space based on the evaluation index, The evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of the multiple biological waveforms over a predetermined period. The waveform comparison unit uses a cross-correlation function to calculate a plurality of cross-correlation coefficients for the plurality of biological waveforms over the predetermined period as the comparison result. The determination unit determines the number of people in the space based on the cross-correlation coefficient calculated by the waveform comparison unit. The plurality of detection signals include a first detection signal and a second detection signal. The waveform estimation unit estimates a first biological waveform corresponding to the first detection signal and a second biological waveform corresponding to the second detection signal. The waveform comparison unit calculates a plurality of cross-correlation coefficients for the first biological waveform and the second biological waveform over the predetermined period as the comparison result. The determination unit, Based on the comparison results and a predetermined threshold, it is determined whether there is one person or two people in the space. The predetermined threshold is a value based on the reference cross-correlation coefficient corresponding to the peak value in the frequency distribution of the multiple reference cross-correlation coefficients, which are calculated using the first person's bio-waveform estimated by the waveform estimation unit when only the first person is present in the space and the second person's bio-waveform estimated by the waveform estimation unit when only the second person is present in the space during the predetermined period. Signal processing system.

6. A signal acquisition unit that acquires multiple detection signals corresponding to each of the multiple detection areas from a sensor system installed in space that detects people in multiple different detection areas, A waveform estimation unit estimates a bio-waveform related to human biological information from each of the aforementioned plurality of detection signals, A waveform comparison unit that compares a plurality of biological waveforms corresponding to each of the plurality of detection signals, The system includes a determination unit that determines the number of people in the space based on the comparison results of the waveform comparison unit, The plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. The waveform comparison unit calculates an evaluation index representing the similarity of the multiple biological waveforms as a result of the comparison, The determination unit determines the number of people in the space based on the evaluation index, The evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of the multiple biological waveforms over a predetermined period. The waveform comparison unit uses a cross-correlation function to calculate a plurality of cross-correlation coefficients for the plurality of biological waveforms over the predetermined period as the comparison result. The determination unit determines the number of people in the space based on the cross-correlation coefficient calculated by the waveform comparison unit. The plurality of detection signals include a first detection signal and a second detection signal. The waveform estimation unit estimates a first biological waveform corresponding to the first detection signal and a second biological waveform corresponding to the second detection signal. The waveform comparison unit calculates a plurality of cross-correlation coefficients for the first biological waveform and the second biological waveform over the predetermined period as the comparison result. The determination unit, Based on the comparison results and a predetermined threshold, it is determined whether there is one person or two people in the space. The waveform estimation unit includes a first waveform estimation unit for estimating the first biological waveform and a second waveform estimation unit for estimating the second biological waveform. The predetermined threshold is a value based on a reference cross-correlation coefficient corresponding to the peak value in a frequency distribution that combines a plurality of first reference cross-correlation coefficients, which are a plurality of reference cross-correlation coefficients for a predetermined period when only the first person is present in the space, and a plurality of second reference cross-correlation coefficients, which are a plurality of reference cross-correlation coefficients for a predetermined period when only the second person is present in the space. The plurality of first reference cross-correlation coefficients are obtained from the first waveform, which is the first biological waveform estimated by the first waveform estimation unit for the predetermined period, and the second waveform, which is the second biological waveform estimated by the second waveform estimation unit for the predetermined period, when only the first person is present in the space. The plurality of second reference cross-correlation coefficients are obtained from the third waveform, which is the first biological waveform estimated by the first waveform estimation unit for the predetermined period, and the fourth waveform, which is the second biological waveform estimated by the second waveform estimation unit for the predetermined period, when only the second person is present in the space. Signal processing system.

7. A signal acquisition unit that acquires multiple detection signals corresponding to each of the multiple detection areas from a sensor system installed in space that detects people in multiple different detection areas, A waveform estimation unit estimates a bio-waveform related to human biological information from each of the aforementioned plurality of detection signals, A waveform comparison unit that compares a plurality of biological waveforms corresponding to each of the plurality of detection signals, A determination unit determines the number of people in the space based on the comparison results of the waveform comparison unit, Waveform adjustment section, It comprises a display processing unit and, The plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. The waveform comparison unit calculates an evaluation index representing the similarity of the multiple biological waveforms as a result of the comparison, The determination unit determines the number of people in the space based on the evaluation index, The evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of the multiple biological waveforms over a predetermined period. The waveform comparison unit uses a cross-correlation function to calculate a plurality of cross-correlation coefficients for the plurality of biological waveforms over the predetermined period as the comparison result. The determination unit determines the number of people in the space based on the cross-correlation coefficient calculated by the waveform comparison unit. The plurality of detection signals include a first detection signal and a second detection signal. The waveform estimation unit estimates a first biological waveform corresponding to the first detection signal and a second biological waveform corresponding to the second detection signal. The waveform comparison unit calculates a plurality of cross-correlation coefficients for the first biological waveform and the second biological waveform over the predetermined period as the comparison result. The determination unit, Based on the comparison results and a predetermined threshold, it is determined whether there is one person or two people in the space. The waveform adjustment unit is If the determination unit determines that one person is present in the space, it calculates the first average amplitude of the first biological waveform and the second average amplitude of the second biological waveform, and adjusts the amplitude of the biological waveform with a smaller average amplitude by multiplying it by a suppression coefficient, while no amplitude adjustment is performed on the biological waveform with a larger average amplitude. If the determination unit determines that two people are present in the space, the amplitude of both the first and second biological waveforms will not be adjusted. The display processing unit, If the determination unit determines that one person is present in the space, it displays on the display unit the first bio-waveform and the second bio-waveform, which have had their amplitudes adjusted, and the bio-waveform with a large average amplitude, which has not had its amplitude adjusted, from among the first and second bio-waveforms. If the determination unit determines that two people are present in the space, it will display the first and second bio-waveforms, which have not undergone amplitude adjustment, on the display unit. Signal processing system.

8. The waveform comparison unit calculates multiple cross-correlation coefficients for each of the multiple predetermined periods, extracts the cross-correlation coefficient corresponding to the peak value in the frequency distribution of the multiple cross-correlation coefficients as the maximum cross-correlation coefficient, and uses the calculated value obtained from the extracted maximum cross-correlation coefficients as the comparison result. The determination unit determines the number of people in the space based on the calculated value. The signal processing system according to any one of claims 5 to 7.

9. The calculated value is one of the mean, median, maximum, and minimum values ​​of the plurality of maximum cross-correlation coefficients. The signal processing system according to claim 8.

10. The waveform estimation unit estimates the plurality of biological waveforms corresponding to the plurality of detection signals using a learning model for data estimation constructed by a recurrent neural network. The signal processing system according to any one of claims 1 to 7.

11. The aforementioned biological waveform is a human respiratory waveform. The signal processing system according to any one of claims 1 to 7.

12. A signal processing system according to any one of claims 1 to 7, The system includes a sensor system for detecting people in the aforementioned space. Human detection system.

13. A signal acquisition step involves acquiring multiple detection signals corresponding to each of the multiple detection areas from a sensor system installed in a space that detects people in multiple different detection areas, A waveform estimation step in which a bio-waveform relating to human bio-information is estimated from each of the aforementioned plurality of detection signals, A waveform comparison step of comparing a plurality of biological waveforms corresponding to each of the plurality of detection signals, The process includes a determination step of determining the number of people in the space based on the comparison results in the waveform comparison step, The plurality of detection signals are signals generated by the sensor system based on an oscillation signal corresponding to a first radio wave output by the sensor system toward a corresponding detection area and a reception signal corresponding to a second radio wave received by the sensor system as a reflected wave in the corresponding detection area, which is the reflected wave of the first radio wave. In the waveform comparison step, an evaluation index representing the similarity of the multiple biological waveforms is calculated as a result of the comparison. In the determination step, the number of people in the space is determined based on the evaluation index. The evaluation index is a cross-correlation coefficient that represents the cross-correlation relationship of the multiple biological waveforms over a predetermined period. In the waveform comparison step, a cross-correlation function is used to calculate a plurality of cross-correlation coefficients for the plurality of biological waveforms over the predetermined period as the comparison result. In the determination step, the number of people in the space is determined based on the cross-correlation coefficient calculated in the waveform comparison step. In the waveform comparison step, for each of the multiple predetermined periods, multiple cross-correlation coefficients for the predetermined period are calculated, and the cross-correlation coefficient corresponding to the peak value in the frequency distribution of the multiple cross-correlation coefficients is extracted as the maximum cross-correlation coefficient, and the calculated value obtained from the extracted multiple maximum cross-correlation coefficients is used as the comparison result. In the determination step, the number of people in the space is determined based on the calculated value. Signal processing method.

14. A program for causing a computer system to execute the signal processing method described in claim 13.

Images