Walking estimation device and walking estimation program

Abstract

To provide a walking estimation device and a walking estimation program capable of estimating a walking mode of a pedestrian and calculating an index value highly correlated with a walking speed.SOLUTION: The walking estimation device 1 includes the acceleration sensor 80 that detects acceleration when the pedestrian 100 walks, the peak detection unit 13 that detects a peak based on waveform information of the acceleration obtained by the acceleration sensor 80, and the index value calculation unit 16 that calculates an index value indicating a walking mode of the pedestrian 100 based on waveform information in which a value of the acceleration is equal to or less than a predetermined acceleration between peaks obtained by the peak detection unit 13.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a walking estimation device and the like that calculates an index value that has a high correlation with walking speed. [Background technology]

[0002] It has been known that walking speed is a useful index for evaluating diseases such as Parkinson's disease that cause gait disorders (such as freezing of gait and lunging) (see, for example, Non-Patent Document 1). Methods for estimating walking speed using measuring devices such as acceleration sensors have been widely studied. As a method for estimating walking speed, for example, Non-Patent Document 2 discloses a method for calculating speed by integrating acceleration in various algorithms, and Non-Patent Document 3 discloses a method for estimating speed from the signal energy of vertical acceleration in a walking frequency band of 1 to 3 Hz, with the aim of evaluating only the vertical vibrations caused by walking. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Masaki Sekine and six others, "Evaluation of Gait Disorders in Parkinson's Disease Patients Using an Accelerometer," Life Support, 2004, Vol. 16, No. 3, pp. 90-97 [Non-patent document 2] A. Soltani et al., "Algorithms for Walking Speed ​​Estimation Using a Lower-Back-Worn Inertial Sensor: A Cross-Validation on Speed ​​Ranges," IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2021, vol. 29, pp. 1955-1964 [Non-patent document 3] Mitsutoshi Suson and five others, "Walking Speed ​​Estimation Using Signal Energy of Acceleration in the Walking Frequency Band," Transactions of the Society of Instrument and Control Engineers, 2004, Vol. 40, No. 4, pp. 371-375 Summary of the Invention [Problem to be solved by the invention]

[0004] The walking speed estimation methods based on acceleration integration and specific frequency domain analysis disclosed in Non-Patent Documents 2 and 3 require conversion (FFT or DFT) of the entire range of signals detected by the acceleration sensor, and are therefore susceptible to the influence of sensor-specific noise and white noise. Therefore, careful calibration and an experimental environment are required to achieve highly accurate speed estimation. Furthermore, in order to cancel out the noise mentioned above, it becomes necessary to calibrate the estimated walking speed using second sensor information such as location information from a GPS, which makes the system more complicated.

[0005] The present invention has been made to solve the above-mentioned problems, and aims to provide a gait estimation device and a gait estimation program that calculate an index value that has a high correlation with walking speed based on information obtained by an acceleration sensor. [Means for solving the problem]

[0006] The gait estimation device according to the present invention includes an acceleration sensor that detects acceleration when a pedestrian walks, peak detection means that detects peaks based on waveform information of acceleration obtained by the acceleration sensor, and index value calculation means that calculates an index value that indicates the pedestrian's walking pattern based on waveform information where the acceleration value is equal to or less than a predetermined acceleration between each of the peaks obtained by the peak detection means.

[0007] In this way, in the present invention, the index value calculation means calculates an index value indicating the walking mode of the pedestrian based on waveform information in which the acceleration value is equal to or less than a predetermined acceleration. Therefore, the index value is calculated using waveform information of acceleration in a range in which the influence of impacts such as the pedestrian touching down is small, using the detection results of only the acceleration sensor, and an index value that is highly correlated with walking speed can be obtained. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a schematic diagram showing a usage mode of the gait estimation device according to this embodiment. [Figure 2] 1 is a hardware configuration diagram of a gait estimation device according to a first embodiment of the present invention. [Figure 3] 1 is a functional block diagram showing the configuration of a gait estimation device according to a first embodiment of the present invention. [Figure 4] 3 is a flowchart showing the operation of the gait estimation device according to the first embodiment of the present invention. [Figure 5] FIG. 2 is a schematic diagram for explaining the operation of the gait estimation device according to the first embodiment of the present invention. [Figure 6] FIG. 6 is a schematic diagram for explaining the operation of a gait estimation device according to a second embodiment of the present invention. [Figure 7] FIG. 6 is a schematic diagram for explaining the operation of a gait estimation device according to a second embodiment of the present invention. [Figure 8] 10 is a flowchart showing the operation of the gait estimation device according to the second embodiment of the present invention. [Figure 9] 10 is a graph showing the correlation between walking speed and index value in an example. DETAILED DESCRIPTION OF THE INVENTION

[0009] (First embodiment of the present invention) A gait estimation device and a gait estimation program according to a first embodiment of the present invention will be described below with reference to Figures 1 to 6. The gait estimation device according to this embodiment is carried by a pedestrian and realizes a function of calculating an index value indicating the gait pattern of the pedestrian based on acceleration waveform information obtained by a built-in acceleration sensor.

[0010] FIG. 1(a) is a schematic diagram showing a usage (carrying) manner of a gait estimation device according to this embodiment, and FIG. 1(b) is a graph showing a change in acceleration of a walker over time. As shown in FIG. 1(a), the gait estimation device 1 according to this embodiment is carried by a walker 100 on his / her hand or waist. One step is defined as a state from when one foot is on the ground (state A) to when the other foot is on the ground (state C). Based on waveform information of acceleration obtained by an acceleration sensor, the device calculates and displays an index value indicating the walking behavior of the walker 100 over a predetermined number of steps. The index value, which is highly correlated with the walking speed of the walker 100, can be used to predict, for example, gait disorders due to mild cognitive impairment (MCI), and can be used for early recovery, treatment, etc.

[0011] Here, one step is defined as a period from state A, where one foot of the walker 100 touches the ground (heel strike, or initial contact), through state B, where the other foot is raised and the knee reaches its highest point, to state C, where the other foot touches the ground (heel strike, or contralateral initial contact). States A to C in one step can be observed on a graph showing acceleration waveform information, with acceleration on the vertical axis and time on the horizontal axis, as shown in FIG. 1(b). States A and C, which correspond to heel strike, exhibit large accelerations that peak in the waveform corresponding to one step because the impact when the walker 100 touches the ground is large, and are detected as adjacent peaks. Furthermore, state B is a state in which the other foot, which is opposite the foot that initially touches the ground, is positioned midway between the process in which the knee leaves the ground surface and moves toward its highest point, and the process in which the knee moves from its highest point toward the ground surface. Therefore, state B exhibits the smallest acceleration in the waveform corresponding to one step and is detected as a valley between the peaks.

[0012] Furthermore, when transitioning from state A to state B, the other leg is raised, causing acceleration of the body in the direction opposite to the direction of gravitational acceleration. Therefore, the acceleration measured by the acceleration sensor gradually decreases from the peak corresponding to state A, and the acceleration reaches its minimum value in state B. When transitioning from state B to state C, the knee of the other leg is lowered from its highest position, causing acceleration of the body in the same direction as the direction of gravitational acceleration. Therefore, the acceleration measured by the acceleration sensor increases from the peak corresponding to state B and finally reaches the peak corresponding to state C.

[0013] The gait estimation device 1 is a mobile terminal such as a smartphone that has a built-in acceleration sensor, and the pedestrian 100 measures the acceleration over a predetermined number of steps by operating operation buttons displayed on the display of the gait estimation device 1. The gait estimation device 1 calculates an index value from the waveform information of the acceleration obtained by the acceleration sensor and displays the index value on the display. The walking estimation device 1 may be a dedicated terminal equipped with at least an acceleration sensor and capable of calculating an index value highly correlated with walking speed, or may be a general-purpose terminal such as a smartphone, tablet, laptop computer, or wearable device.

[0014] 2 is a hardware configuration diagram of a gait estimation device according to this embodiment. The gait estimation device 1 includes, as its main components, a processor 10 such as a CPU that executes a program, a ROM 20 that stores data in a non-volatile manner, a RAM 30 that includes a work area for the processor 10 when the program is executed and an area for volatilely storing data generated by program execution or input data, an HDD 40 that stores programs and data in a non-volatile manner, a communication I / F 50, an input device 60 that accepts operations on the gait estimation device 1, a display 70 that displays various information, and a three-axis acceleration sensor 80 that detects acceleration and tilt in each of three axes (X-axis, Y-axis, and Z-axis), but is not limited thereto. The components are interconnected by a data bus.

[0015] The communication I / F 50 is an interface for communication between devices. The input device 60 is an operation key configured by arranging touch sensors and button switches etc. that are stacked on the display screen of the display 70, and is used to input various operation signals. The acceleration sensor 80 measures acceleration, which is a measured value of a physical quantity that reflects the walking manner of the pedestrian 100.

[0016] 3 is a functional block diagram showing the configuration of a gait estimation device according to this embodiment. The gait estimation device 1 includes an acceleration sensor 80 that detects accelerations in the X-axis, Y-axis, and Z-axis directions when a pedestrian 100 walks, a calculation unit 11 that combines the accelerations in the three axes detected by the acceleration sensor 80 and calculates an index value indicating the walking pattern of the pedestrian 100 from waveform information of the combined acceleration, a waveform information storage unit 21 that stores the waveform information of the acceleration combined by the calculation unit 11, and a display 70 that displays the index value calculated by the calculation unit 11.

[0017] 3 is realized by causing a computer processor to function as each processing unit by a gait estimation program. The waveform information storage unit 21 is a data unit stored in an internal storage such as a ROM or HDD of the computer, but may also be stored online in an external storage. Furthermore, a network may be constructed so that data is stored on a cloud server.

[0018] The calculation unit 11 includes an acceleration synthesis unit 12 that synthesizes three-axis acceleration components detected by the acceleration sensor 80 and stores the synthesized data in a waveform information storage unit 21; a peak detection unit 13 that detects peaks of the acceleration waveform based on the acceleration waveform information stored in the waveform information storage unit 21; an acceleration information acquisition unit 14 that acquires acceleration information using the minimum acceleration value between peaks based on the acceleration waveform information; a progress information acquisition unit 15 that acquires progress information including at least time-dependent elements between peaks based on the acceleration waveform information; an index value calculation unit 16 that calculates an index value indicating the walking pattern of the pedestrian 100 based on the acceleration information acquired by the acceleration information acquisition unit 14 and the progress information acquired by the progress information acquisition unit 15; and a display control unit 17 that displays the index value calculated by the index value calculation unit 16 on a display 70.

[0019] The peak detection unit 13 detects peaks from the acceleration waveform information stored in the waveform information storage unit 21. A known method for detecting peaks, such as differentiating the acceleration waveform information, can be used. However, for example, the acceleration waveform can be divided into predetermined time intervals, and the maximum acceleration within each interval that exceeds a preset threshold can be determined as a peak. Alternatively, the average acceleration value or a preset threshold can be used as a reference acceleration, and the point at which the acceleration rises above the reference acceleration (transitions from a value smaller than the reference acceleration to a value larger than the reference acceleration) can be used as a reference point. The range from any reference point to the adjacent reference point can be treated as a single wave, and the point showing the maximum acceleration within the wave can be determined as a peak.

[0020] The acceleration information acquisition unit 14 acquires acceleration information using the minimum acceleration value between the peaks based on the peaks detected by the peak detection unit 13. The acceleration information using the minimum acceleration value may be the minimum acceleration value itself, or may be the difference in acceleration between a predetermined acceleration as a reference and the minimum acceleration value.

[0021] The progress information acquiring unit 15 acquires progress information including at least a time-dependent element between peaks based on the peaks detected by the peak detecting unit 13. The progress information may be the number of acceleration data points between peaks below a predetermined acceleration, or the elapsed time. The number of data points or the elapsed time may also be the range from any adjacent peak to the minimum acceleration value.

[0022] The index value calculation unit 16 calculates an index value based on waveform information of accelerations that are equal to or less than a predetermined acceleration, among the waveform information of accelerations synthesized by the acceleration synthesis unit 12. Specifically, the index value calculation unit 16 calculates the index value based on the acceleration information acquired by the acceleration information acquisition unit 14 and the progress information acquired by the progress information acquisition unit 15. For example, the index value can be calculated by dividing the average value of the acceleration information for all steps by the average value of the progress information, as shown in the following formula (1).

[0023] Index value = average value of acceleration information / average value of progress information (1)

[0024] Here, the predetermined acceleration is not particularly limited as long as it is an acceleration that is at least smaller than the peak acceleration (peak value) detected by the peak detection unit 13. For example, it may be a value between the smallest peak value detected by the peak detection unit 13 and the largest minimum acceleration value between peaks acquired by the acceleration information acquisition unit 14, or it may be a value between the gravitational acceleration (9.8 m / s) detected by the acceleration sensor 80 when the pedestrian 100 is stopped. 2 ) can also be used.

[0025] Note that the index value calculation unit 16 does not necessarily need to use all waveform information that is below a predetermined acceleration as the target for calculating the index value; it can also use only a portion of the waveform, for example, the acceleration waveform in the first half of the minimum acceleration value as the boundary, for calculating the index value.

[0026] The display control unit 17 displays the index value on the display 70. The display control unit 17 may display a value such as a score derived from the index value on the display 70 instead of the index value.

[0027] The walking estimation device 1 configured as described above may constantly measure the acceleration according to the walking of the pedestrian 100 and calculate the index value, or may measure the acceleration for a predetermined walking time, walking distance, etc. and calculate the index value by operation of the pedestrian 100.

[0028] Next, the operation of the gait estimation device according to this embodiment will be described. Fig. 4 is a flowchart showing the operation of the gait estimation device according to this embodiment, and Figs. 5 and 6 are schematic diagrams for explaining the operation of the gait estimation device according to this embodiment. The following description will be given taking the acceleration waveform information shown in Figs. 5 and 6 as an example, but is not limited to this.

[0029] 5, when acceleration sensor 80 detects acceleration over time in the three axes of X-axis, Y-axis, and Z-axis corresponding to the walking of walker 100, acceleration synthesis unit 12 synthesizes the acceleration components of these three axes (S10). Waveform information of the synthesized acceleration is stored in waveform information storage unit 21.

[0030] The peak detection unit 13 detects a peak corresponding to a heel strike in the walking of the walker 100 based on the waveform information of the acceleration synthesized by the acceleration synthesis unit 12 (S11). At this time, the peak detection unit 13 detects a peak corresponding to a heel strike in the walking of the walker 100 based on the acceleration threshold aT1 (for example, aT1=11 m / s 2 ) or more are detected as peaks corresponding to heel strikes. Specifically, seven peaks P1 to P7 are detected by peak detection unit 13, and six steps of pedestrian 100 are used to calculate an index value indicating the walking style of pedestrian 100.

[0031] 6, the acceleration information acquiring unit 14 acquires an acceleration difference (Δan) between a predetermined acceleration aT2 between adjacent peaks and the minimum acceleration value between the peaks based on the peaks detected by the peak detecting unit 13 (S12). Specifically, the acceleration information acquiring unit 14 acquires acceleration differences Δa1 to Δa6 between P1 and P2, between P2 and P3, ..., and between P6 and P7. In FIG. 6, the predetermined acceleration aT2 is shown to be a value smaller than the acceleration threshold aT1 set in the peak detection unit 13, but this is not limited thereto, and the acceleration aT2 may be equal to or greater than the acceleration threshold aT1.

[0032] Based on the peaks detected by the peak detection unit 13, the progress information acquisition unit 15 acquires the number of acceleration data Nn between peaks as progress information (S13). As in S12, the number of acceleration data is acquired using a predetermined acceleration aT2 as a threshold, and refers to the number of acceleration samples that are equal to or less than the acceleration threshold aT2. Specifically, data numbers N1 to N6 are acquired corresponding to the intervals P1-P2, P2-P3, ..., and P6-P7, respectively. The acceleration data is acquired by the acceleration sensor 80, for example, every 20 ms.

[0033] The index value calculation unit 16 calculates an index value indicating the walking pattern of the pedestrian 100 from the average value of the acceleration difference (Δan) acquired by the acceleration information acquisition unit 14 and the average value of the number of acceleration data points (Nn) between each peak acquired by the progress information acquisition unit 15 (S14), and then terminates the processing. Specifically, as shown in the following formula (2), the index value is calculated by dividing the average value of the differences Δa1 to Δa6 between the peaks by the average value of the numbers N1 to N6 of acceleration data between the peaks.

[0034] Index value = average value of acceleration difference / average value of number of data (2) where: Average value of acceleration difference = (Δa1 + Δa2 + + Δa6) / 6 Average number of data points = (N1 + N2 + + N6) / 6

[0035] The index value may also be calculated by dividing the sum of the minimum acceleration values ​​by the sum of the number of data points, without separately calculating the average value of the acceleration differences and the average value of the number of data points. The index value calculated in this way is the same as the index value calculated by dividing the average value of the acceleration differences by the average value of the number of data points.

[0036] As described above, the device is equipped with acceleration sensor 80 that detects the acceleration of pedestrian 100 when walking, peak detection unit 13 that detects peaks based on waveform information of the acceleration obtained by acceleration sensor 80, and index value calculation unit 16 that calculates an index value that indicates the walking mode of pedestrian 100 based on waveform information where the acceleration value is below a predetermined acceleration between each of the peaks obtained by peak detection unit 13.Therefore, using the detection results of acceleration sensor 80 alone, the index value is calculated using waveform information of acceleration in a range that is less affected by impacts such as pedestrian 100 touching the ground, and an index value that is highly correlated with walking speed can be obtained.

[0037] The device also includes an acceleration information acquisition unit 14 that acquires acceleration information using the minimum acceleration value between peaks based on waveform information, and a progress information acquisition unit 15 that acquires progress information including at least time-varying elements between peaks based on waveform information. The index value calculation unit 16 calculates an index value based on the acceleration information and progress information, so that the index value is calculated using two variables, acceleration information and progress information, which can be easily detected by the acceleration sensor 80, as indexes. This makes it possible to calculate an index value that is highly correlated with walking speed with a simple configuration and processing without complicating the device.

[0038] Furthermore, because the acceleration information is the difference in acceleration between a predetermined acceleration and the minimum acceleration value between peaks, the index value is calculated using acceleration information that indicates the amount of change in acceleration and progress information that indicates the amount of change in an element over time, making it possible to reflect the movement of the walker 100 in the index value. Furthermore, by calculating the index value from part of the waveform information below the predetermined acceleration, for example, waveform information corresponding to the movement from when the foot of the walker 100 leaves the ground until the knee rises to the highest position, it becomes possible to evaluate the strength of the muscle strength in each leg of the walker 100 and the smoothness of the joint movement.

[0039] Furthermore, since the index value calculation unit 16 calculates the index value by dividing the average value of the acceleration information by the average value of the progress information, there is no need for complex calculation processes such as conversion and integration of the acceleration signal, which reduces the amount of calculation by the calculation unit 11 and suppresses power consumption.

[0040] Furthermore, since the specified acceleration is the gravitational acceleration, the acceleration in a stationary state is used as the reference, and the index value can be calculated based on waveform information of movements with movement relative to the reference acceleration.

[0041] In addition, after the peak detection unit 13 detects peaks from the acceleration waveform information, the walking estimation device 1 can also calculate index values ​​only between consecutive peaks where the progress information including the time-dependent elements between the peaks is within a predetermined range. For example, when the gait estimation device 1 constantly monitors the pedestrian 100, it also monitors daily movements other than walking. If an index value is calculated including such irregular daily movements, the discrepancy in correlation between the index value and walking speed will become large. Therefore, by targeting only regular movements, i.e., cases where progress information is continuously within a predetermined numerical range, it is possible to calculate an index value that is highly correlated with walking speed even when the pedestrian 100 is constantly monitored.

[0042] Furthermore, the index value calculation unit 16 can also be configured to calculate the index value by integrating the range in which the acceleration between these peaks is equal to or less than a predetermined acceleration after the peak detection unit 13 detects peaks based on the waveform information of the acceleration. In this embodiment, when calculating the index value, waveform information of the acceleration that is equal to or less than the predetermined acceleration is targeted. Therefore, even when the index value is calculated by integration, it is possible to calculate an index value that is highly correlated with walking speed while suppressing the influence of noise, etc.

[0043] (Second embodiment of the present invention) A gait estimation device and a gait estimation program according to a second embodiment of the present invention will be described with reference to Figures 7 and 8. In the gait estimation device 1 according to this embodiment, the peak detection unit 13 detects peaks from waveform information of acceleration, determines whether or not each of the detected intervals between adjacent peaks corresponds to one actual step of the pedestrian 100, and if it determines that the interval does not correspond to one step, calculates an index value by processing only the intervals between peaks, excluding the waveform between the peaks. In this embodiment, explanations that overlap with those in the first embodiment will be omitted.

[0044] The operation of the gait estimation device according to this embodiment will be described below. Fig. 7 is a schematic diagram for explaining the operation of the gait estimation device according to this embodiment, and Fig. 8 is a flowchart showing the operation of the gait estimation device according to this embodiment. The following description will be given taking the acceleration waveform information shown in Fig. 7 as an example, but is not limited to this.

[0045] 4 in the first embodiment, so a detailed description will be omitted, but in S10, the peak detection unit 13 detects only peaks P1 to P4, peaks P6, and peaks P7, where the acceleration is equal to or greater than the acceleration threshold aT1, from the waveform information of the acceleration synthesized by the acceleration synthesis unit 12, as peaks corresponding to heel strikes, and sets these as target peaks to be used in each process from S10 onwards. Peak P5 has an acceleration value smaller than the acceleration threshold aT1, so it is excluded from the target peaks in each process from S10 onwards.

[0046] Next, the peak detection unit 13 acquires the elapsed time tn between each peak of the pedestrian 100, and determines whether the elapsed time tn is equal to or less than a predetermined elapsed time T (tn≦T) (S20). Here, the elapsed time T can be, for example, 1.5×t0≦T≦2.0×t0, where t0 is the average time it takes for a typical pedestrian to take one step. This average time t0 can be changed as appropriate depending on the attributes of the pedestrian 100, such as the age, sex, and height. If the elapsed time t1 between the first peaks (between P1 and P2) is equal to or less than the predetermined elapsed time T (S20: YES), it is then determined whether or not the peak interval (between P1 and P2) is the last peak interval (S21). If the interval between P1 and P2 is not the last peak interval in the waveform information of the acceleration synthesized by the acceleration synthesis unit 12 (S21: NO), the elapsed time t2 between the next peak interval (between P2 and P3) is obtained (S22), and the process returns to S20 to repeat the same process. On the other hand, in S20, when the elapsed time tn between peaks is greater than a predetermined elapsed time T (S20: NO), the peak detection unit 13 excludes the period between the peaks from the processing target (S23). Specifically, when it is determined that the elapsed time t4 between P4 and P6 obtained after the determination of the elapsed time t3 between P3 and P4 is greater than the predetermined elapsed time T (t4 > T, S20: NO), the period between P4 and P6 is excluded from the target for calculating the index value, and the acceleration difference (Δa4) and the number of data (N4) are not acquired. Then, it is determined whether the period between P4 and P6 for which it has been determined that tn < T is the last period between peaks (S21). If it is not the last period between peaks (S21: NO), the elapsed time t5 of the next period between peaks (between P7 and P8) is acquired (S22), and the process returns to S20 to repeat the same process.

[0047] In S21, when the period between peaks for which it has been determined that tn ≤ T is the last period between peaks (between P6 and P7) (S21: YES), the acceleration information acquisition unit 14 acquires the acceleration differences (Δan) for the periods between peaks that are the targets detected by the peak detection unit 13. Specifically, the acceleration differences between the predetermined acceleration aT2 and the minimum acceleration value between peaks are acquired for between P1 and P2, between P2 and P3, between P3 and P4, and between P6 and P7 (S24).

[0048] The elapsed information acquisition unit 15 acquires the number of acceleration data for the periods between peaks that are the targets detected by the peak detection unit 13. Specifically, the number of acceleration data is acquired from the waveform information below the predetermined acceleration aT2 for between P1 and P2, between P2 and P3, between P3 and P4, and between P6 and P7 (S25).

[0049] The index value calculation unit 16 calculates the index value based on the above formula (2) from the average value of the acceleration differences (Δan) acquired by the acceleration information acquisition unit 14 and the average value of the number of data (Nn) acquired by the elapsed information acquisition unit 15 (S26), and ends the process.

[0050] As described above, when the time between peaks is equal to or longer than a predetermined time, the index value calculation unit 16 calculates the index value by excluding the waveform between the peaks. Therefore, for example, peaks with data counts of two or more steps can be excluded from the target for calculating the index value, and an index value having a higher correlation with walking speed can be calculated.

[0051] In addition, if the minimum acceleration value between peaks acquired by the acceleration information acquisition unit 14 is greater than a predetermined acceleration aT2, the index value calculation unit 16 may exclude the peak from the processing target and calculate the index value, as described above.

[0052] In addition, when there is a peak interval that has been excluded when calculating the index value, particularly when there is a peak interval that has been excluded between two peaks that were the subject of the index value calculation, the display control unit 17 may display the reason for the exclusion (insufficient peak value, insufficient minimum value, excessive elapsed time between peaks, etc.) on the display 70 together with the index value.

[0053] The above-described embodiments can be implemented in appropriate combinations. [Example]

[0054] The index value of the present invention was verified as follows: The index value was calculated by dividing the average value of the acceleration difference between the gravitational acceleration and the minimum acceleration value between the peaks by the average value of the number of data points, using waveform information of the obtained acceleration that is equal to or less than the gravitational acceleration.

[0055] (1) Verification 1 At the "Sasaguri Genkimon" physical examination held in Sasaguri Town, Kasuya District, Fukuoka Prefecture from April 25, 2022 to January 31, 2023, a 10-meter walking experiment was conducted on elderly people in their 60s to 90s using the "AYUMI EYE" walking accelerometer (manufactured by Waseda Elderly Health Foundation, Inc.), and acceleration data of normal walking was collected from a total of 353 subjects. Index values ​​were calculated from this walking data group. Separately, average speed was calculated from the walking distance (10 meters) and the time required for walking, and this was used as the reference average walking speed. The results are shown in Figure 9. Figure 9 is a graph showing the correlation between walking speed and index value, with walking speed (m / s) on the horizontal axis and index value on the vertical axis. The correlation coefficient between the index value and the reference average walking speed was 0.803, indicating a high correlation.

[0056] (2) Verification 2 In a walking measurement experiment conducted at Seiko Electric Manufacturing Co., Ltd. and Fukuoka Institute of Technology from January to February 2024, a total of nine employees and students in their 20s and 30s were asked to walk 10 meters in three patterns: normal walking, brisk walking, and slow walking using an Android (registered trademark) smartphone equipped with an acceleration sensor, and a total of 81 walking acceleration data were collected. An index value was calculated for this walking data group. Separately, the average walking speed was calculated from the walking distance (10 m) and the time required for walking, and this was used as the reference average walking speed. The correlation coefficient between the index value and the reference average walking speed was 0.794, indicating a high correlation. [Explanation of symbols]

[0057] 1. Gait estimation device 10 processors 11 Arithmetic section 12 Acceleration synthesis section 13 Peak detector 14 Acceleration information acquisition section 15 Progress information acquisition unit 16 Index value calculation unit 17 Display control unit 20 ROM 21 Waveform information storage section 30 RAM 40 HDD 50 Communication I / F 60 Input Device 70 Display 80 Acceleration Sensor 100 pedestrians

Claims

1. an acceleration sensor that detects acceleration when a pedestrian walks; a peak detection means for detecting a peak based on waveform information of acceleration obtained by the acceleration sensor; and index value calculation means for calculating an index value indicating the walking pattern of the pedestrian based on waveform information in which the acceleration value is equal to or less than a predetermined acceleration between each of the peaks obtained by the peak detection means.

2. The gait estimation device according to claim 1, an acceleration information acquiring means for acquiring acceleration information using the minimum acceleration value between the peaks based on the waveform information; a progress information acquiring means for acquiring progress information including at least a time-dependent element between the peaks based on the waveform information, The gait estimation device is characterized in that the index value calculation means calculates the index value based on the acceleration information and the progress information.

3. The gait estimation device according to claim 2, The gait estimation device is characterized in that the acceleration information is an acceleration difference between the predetermined acceleration and the minimum acceleration value between the peaks.

4. The gait estimation device according to claim 3, The gait estimation device, wherein the index value calculation means calculates the index value by dividing an average value of the acceleration information by an average value of the progress information.

5. The gait estimation device according to claim 1, The gait estimation device, wherein the predetermined acceleration is gravitational acceleration.

6. The gait estimation device according to claim 1, The gait estimation device is characterized in that, when the time between the peaks is equal to or longer than a predetermined time, the index value calculation means calculates the index value by excluding the waveform between the peaks.

7. a peak detection means for detecting a peak based on waveform information of acceleration obtained by an acceleration sensor that detects acceleration when a pedestrian walks; A gait estimation program that causes a computer to function as index value calculation means that calculates an index value that indicates the walking pattern of the pedestrian, based on waveform information in which the acceleration value is equal to or less than a predetermined acceleration between each of the peaks obtained by the peak detection means.

Images