Measuring device, measuring method, and program

Abstract

Improve the accuracy of step counting. [Solution] The data acquisition unit 101 acquires 3-axis acceleration data for 3 axes output by a 3-axis acceleration sensor attached to the user. The step count estimation unit 103 estimates the user's step count based on the composite acceleration data obtained from the 3-axis acceleration data. The arm swing count estimation unit 104 estimates the user's arm swing count based on specific axis acceleration data from the 3-axis acceleration data that corresponds to the direction in which the user's arm is swung. The step count measurement unit 107 measures the user's step count by switching between the user's step count estimated by the step count estimation unit 103 and the user's arm swing count estimated by the arm swing count estimation unit 104, according to a predetermined standard.

Description

Technical Field

[0001] The present disclosure relates to a measuring device, a measuring method, and a program.

Background Art

[0002] Currently, a measuring device that measures the number of steps taken by a user while walking or running based on the output of an acceleration sensor worn by the user is known. For example, Patent Document 1 describes a measuring device that measures one step for a partial waveform when the time width of the partial waveform having a positive acceleration in the acceleration waveform falls within a reference range.

[0003] In the measuring device described in Patent Document 1, the reference range is adjusted according to the state of the user in order to improve the measurement accuracy. Specifically, in the measuring device described in Patent Document 1, when the state of the user is a running state, a small value is set as the lower limit value of the reference range, and when the state of the user is a walking state, a large value is set as the lower limit value of the reference range.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, as described in Patent Document 1, sufficient measurement accuracy may not be ensured only by adjusting the above-described reference range according to the state of the user. For example, when the user is swinging the arm largely, noise caused by the user's arm swing is included in the acceleration waveform, and it may not be possible to appropriately measure the number of steps from the acceleration waveform. Therefore, a technique for improving the accuracy of step counting is desired.

[0006] This disclosure has been made in view of the above-mentioned problems and aims to provide a measuring device, a measuring method, and a program that improve the accuracy of step counting. [Means for solving the problem]

[0007] To achieve the above objectives, the measuring device relating to this disclosure is A data acquisition means for acquiring 3-axis acceleration data for 3 axes output by a 3-axis acceleration sensor attached to the user, A step count estimation means that estimates the user's step count based on composite acceleration data obtained from the three-axis acceleration data, An arm swing count estimation means estimates the number of times the user swings their arm based on specific axis acceleration data from the three-axis acceleration data that corresponds to the direction in which the user's arm is swung, The system includes a step counting means that measures the user's steps by switching between the user's step count estimated by the step count estimation means and the user's arm swing count estimated by the arm swing count estimation means, according to predetermined criteria. [Effects of the Invention]

[0008] According to this disclosure, the accuracy of step counting can be improved. [Brief explanation of the drawing]

[0009] [Figure 1] Configuration diagram of the measuring device according to the embodiment [Figure 2] A diagram showing the measuring device according to the embodiment being worn by a user. [Figure 3] Functional configuration diagram of the measuring device according to the embodiment [Figure 4] The figures show acceleration waveforms when specific noise is absent. (A) shows the composite acceleration waveform when specific noise is absent, and (B) shows the U-axis acceleration waveform when specific noise is absent. [Figure 5]These figures show acceleration waveforms when specific noise occurs; (A) shows the composite acceleration waveform when specific noise occurs, and (B) shows the U-axis acceleration waveform when specific noise occurs. [Figure 6] A diagram showing a portion of the composite acceleration waveform after offset processing. [Figure 7] A flowchart showing the step counting process performed by the measuring device according to the embodiment. [Figure 8] Figure 7 shows a flowchart illustrating the step count estimation process. [Figure 9] Figure 7 shows a flowchart illustrating the arm swing count estimation process. [Figure 10] Figure showing the measurement results of the measurement error. [Modes for carrying out the invention]

[0010] The embodiments of this disclosure will be described in detail below with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals.

[0011] (Embodiment) Figure 1 shows the configuration of a measuring device 100 according to an embodiment. The measuring device 100 is a device that measures the number of steps taken by a user walking or running. The measuring device 100 is worn on the user's arm and measures the number of steps by detecting the acceleration applied to the measuring device 100. The measuring device 100 can be worn on the finger, wrist, upper arm, etc. Possible measuring devices 100 include a pedometer specifically designed for measuring steps, a smartwatch with a step-counting function, etc. As shown in Figure 1, the measuring device 100 comprises a control unit 11, a storage unit 12, a display unit 13, an operation reception unit 14, a communication unit 15, and a detection unit 16.

[0012] The control unit 11 includes a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), a RTC (Real Time Clock), etc. The CPU, also called a central processing unit, a central arithmetic unit, a processor, a microprocessor, a microcomputer, a DSP (Digital Signal Processor), etc., functions as a central arithmetic processing unit that executes processing and arithmetic operations related to the control of the measuring device 100. In the control unit 11, the CPU reads programs and data stored in the ROM and uses the RAM as a work area to comprehensively control the measuring device 100. The RTC is, for example, an integrated circuit having a timekeeping function. Note that the CPU can specify the current date and time from the time information read from the RTC.

[0013] The storage unit 12 includes a non-volatile semiconductor memory such as a flash memory, an EPROM (Erasable Programmable Read Only Memory), an EEPROM (Electrically Erasable Programmable ROM), etc., and plays a role as a so-called auxiliary storage device. The storage unit 12 stores programs and data used by the control unit 11 to execute various processes. Further, the storage unit 12 stores data generated or acquired when the control unit 11 executes various processes.

[0014] The display unit 13 displays various images according to the control by the control unit 11. For example, the display unit 13 displays a screen for receiving various operations from the user. The display unit 13 includes a touch screen, a liquid crystal display, etc. The operation reception unit 14 receives various operations from the user and supplies information indicating the content of the received operation to the control unit 11. The operation reception unit 14 includes a touch screen, buttons, levers, etc.

[0015] In accordance with the control by the control unit 11, the communication unit 15 communicates with various devices in accordance with well-known wired communication standards or well-known wireless communication standards. Well-known wired communication standards include USB (Universal Serial Bus, registered trademark), Thunderbolt (registered trademark), etc. Well-known wireless communication standards include Wi-Fi (registered trademark), Bluetooth (registered trademark), Zigbee (registered trademark), etc. The communication unit 15 is provided with a communication interface compliant with various communication standards.

[0016] The detection unit 16 detects the acceleration applied to the measuring device 100. The detection unit 16 includes, for example, a three-axis acceleration sensor 161 that detects the acceleration in the U-axis direction, the acceleration in the V-axis direction orthogonal to the U-axis direction, and the acceleration in the W-axis direction orthogonal to the U-axis direction and the V-axis direction. The detection unit 16 outputs three-axis acceleration data, which is time-series data of acceleration for three axes. The three-axis acceleration data includes U-axis acceleration data, which is time-series data of acceleration in the U-axis direction, V-axis acceleration data, which is time-series data of acceleration in the V-axis direction, and W-axis acceleration data, which is time-series data of acceleration in the W-axis direction.

[0017] As shown in FIG. 2, in the present embodiment, the measuring device 100 is a wristwatch-type device worn on the user's wrist. In the present embodiment, when the measuring device 100 is regarded as an analog watch, the U-axis direction is the direction from the position indicating 6 o'clock to the position indicating 12 o'clock, the V-axis direction is the direction from the position indicating 9 o'clock to the position indicating 3 o'clock, and the W-axis direction is the direction from the contact surface to the wrist to the display surface.

[0018] Also, the direction in which the user moves is defined as the X-axis direction, the direction from the position of the right hand to the position of the left hand is defined as the Y-axis direction, and the vertically upward direction is defined as the Z-axis direction. In this case, the X-axis direction, the Y-axis direction, and the Z-axis direction are orthogonal to each other. Here, when the measuring device 100 is disposed on the back side of the user's left wrist, the positive direction of the U-axis is approximately the negative direction of the X-axis, the positive direction of the V-axis is approximately the negative direction of the Z-axis, and the positive direction of the W-axis is approximately the positive direction of the Y-axis.

[0019] Next, the functions of the measuring device 100 will be described with reference to Figure 3. Functionally, the measuring device 100 comprises a data acquisition unit 101, a data processing unit 102, a step count estimation unit 103, an arm swing count estimation unit 104, a walking detection unit 105, a noise generation detection unit 106, a step count measurement unit 107, and a display control unit 108. Each of these functions is realized by software, firmware, or a combination of software and firmware. The software and firmware are written as programs and stored in the ROM or storage unit 12. The CPU then realizes each of these functions by executing the programs stored in the ROM or storage unit 12.

[0020] The data acquisition unit 101 acquires 3-axis acceleration data output by a 3-axis acceleration sensor 161 attached to the user's arm. For example, the data acquisition unit 101 acquires 3-axis acceleration data, which is time-series data of acceleration for all three axes, from the 3-axis acceleration sensor 161 provided by the detection unit 16. In this embodiment, the sampling period for the 3-axis acceleration data is 32 Hz. Therefore, 3-axis acceleration data for one second consists of 32 × 3 = 96 data points. The data acquisition unit 101 is an example of a data acquisition means.

[0021] The data processing unit 102 performs various processes on the 3-axis acceleration data acquired by the data acquisition unit 101, and on various acceleration data obtained from this 3-axis acceleration data. For example, the data processing unit 102 performs synthesis processing, filtering processing, offset processing, extraction processing, etc.

[0022] The synthesis process combines the acceleration data for all three axes. For example, the synthesis process calculates the square root of the sum of the squares of the accelerations in the U-axis direction, V-axis direction, and W-axis direction for each time point, and then obtains the composite acceleration by taking the square root of the calculated sum of squares. By performing the synthesis process on the three-axis acceleration data, composite acceleration data is obtained.

[0023] Filtering is a process that removes high-frequency components from composite acceleration, single-axis acceleration, etc. Generally, the pitch limit during human walking or running is around 5 Hz. Therefore, in filtering, signal components above 5 Hz are removed as noise components. For example, if the sampling frequency of acceleration data is 32 Hz, high-frequency components are removed by a low-pass filter with 7 taps and tap coefficients of 1, 2, 3, 4, 3, 2, 1. By filtering composite acceleration data, single-axis acceleration data, etc., acceleration data from which high-frequency components have been removed is obtained.

[0024] Offset processing is a process that removes the DC component from composite acceleration, acceleration in a single axis direction, etc. When determining the number of steps from changes in acceleration due to walking or running, or when determining the number of arm swings from changes in acceleration due to the user's arm swing, the DC component of acceleration is unnecessary. Therefore, in offset processing, the DC component of acceleration is removed as an offset component. For example, if the sampling frequency of acceleration data is 32Hz, the DC component of acceleration is calculated using an average filter with 17 taps and a tap coefficient of 1. Then, the calculated DC component is removed from the acceleration at each time point. By performing offset processing on composite acceleration data, acceleration data in a single axis direction, etc., acceleration data from which the DC component has been removed is calculated.

[0025] The extraction process involves extracting acceleration data for one axis from three-axis acceleration data. This extraction process on the three-axis acceleration data yields either U-axis acceleration data, V-axis acceleration data, or W-axis acceleration data.

[0026] The step count estimation unit 103 estimates the user's step count during the estimation period based on composite acceleration data obtained from 3-axis acceleration data. As described above, the composite acceleration data is time-series data of composite acceleration obtained by combining the accelerations of the three axes. The estimation period is the period for which the step count and arm swing count are estimated. In this embodiment, the estimation period is the most recent 1 second, and the step count and arm swing count are estimated every second. The step count estimation unit 103 is an example of a step count estimation means.

[0027] The arm swing count estimation unit 104 estimates the number of times the user swings their arm during the estimation period based on specific axis acceleration data from the three-axis acceleration data. Specific axis acceleration data is time-series data of acceleration on a specific axis corresponding to the direction in which the user's arm is swung. In this embodiment, the specific axis is the U-axis, and the specific axis acceleration data is U-axis acceleration data. The arm swing count estimation unit 104 is an example of an arm swing count estimation means.

[0028] The walking detection unit 105 determines whether the user is walking during the estimated period based on the composite acceleration data. For example, the walking detection unit 105 determines whether the user is walking based on the average value of the composite acceleration, the difference between the maximum value and the minimum value of the composite acceleration, etc. The walking detection unit 105 is an example of a walking detection means.

[0029] The noise generation discrimination unit 106 determines whether or not specific noise is occurring during the estimated period based on the composite acceleration data. Specific noise is noise estimated to be caused by the user's arm movements. For example, the noise generation discrimination unit 106 determines whether or not specific noise is occurring based on the number of maximum points in the composite acceleration over a certain period. The noise generation discrimination unit 106 is an example of a noise generation discrimination means.

[0030] The step counting unit 107 measures the number of steps taken by the user while walking or running. The step counting unit 107 determines the number of steps taken during the estimated period and adds the determined number of steps to the cumulative number of steps taken from the start of measurement to the start of the estimated period. In this embodiment, the step counting unit 107 performs the process of determining the number of steps per second and the process of adding the determined number of steps to the cumulative number of steps every second. The step counting unit 107 adopts the estimated number of steps for the estimated period, or twice the estimated number of arm swings for the estimated period, as the number of steps taken during the estimated period.

[0031] Specifically, if the step counting unit 107 determines that the user is not walking during the estimated period, or if it determines that no specific noise is occurring during the estimated period, it adopts the number of steps estimated by the step counting unit 103 as the number of steps taken by the user during the estimated period. On the other hand, if the step counting unit 107 determines that the user is walking during the estimated period, and if it determines that no specific noise is occurring during the estimated period, it adopts twice the number of arm swings estimated by the arm swing counting unit 104 as the number of steps taken by the user during the estimated period.

[0032] Thus, the step counting unit 107 adopts the number of steps estimated based on the composite acceleration data when the user is not walking or when specific noise is not occurring. On the other hand, when the user is walking and specific noise is occurring, the step counting unit 107 adopts twice the number of arm swings estimated based on the specific axis acceleration data. The step counting unit 107 is an example of a step counting means.

[0033] The display control unit 108 displays the measurement results from the step counting unit 107. For example, the display control unit 108 controls the display unit 13 to display the cumulative number of steps calculated by the step counting unit 107. The cumulative number of steps is the cumulative value of the number of steps measured from the start of measurement to the present time.

[0034] The following explains, with reference to Figures 4 and 5, why both estimation of the number of steps using composite acceleration data and estimation of the number of arm swings using specific axis acceleration data are performed.

[0035] Figure 4 shows the acceleration waveform when no specific noise is present. More specifically, Figure 4(A) shows the composite acceleration waveform corresponding to the composite acceleration data after filtering, obtained when no specific noise is present. Figure 4(B) shows the U-axis acceleration waveform corresponding to the U-axis acceleration data after filtering, obtained when no specific noise is present.

[0036] Figure 5 shows the acceleration waveform when a specific noise occurs. More specifically, Figure 5(A) shows the composite acceleration waveform corresponding to the composite acceleration data after filtering acquired when a specific noise occurs. Figure 5(B) shows the U-axis acceleration waveform corresponding to the U-axis acceleration data after filtering acquired when a specific noise occurs. In Figures 4 and 5, the horizontal axis is the time axis, and the vertical axis is the axis representing various accelerations.

[0037] As shown in Figure 4(A), when no specific noise is present, the composite acceleration waveform clearly shows waves with periods corresponding to walking or running pitch. The number of these waves corresponds to the number of steps. Therefore, when no specific noise is present, the number of steps can be accurately estimated from the composite acceleration data.

[0038] As shown in Figure 4(B), when no specific noise is present, the U-axis acceleration waveform does not show a wave with a period corresponding to the arm swing pitch. Therefore, when no specific noise is present, it is difficult to estimate the number of arm swings from the U-axis acceleration data.

[0039] On the other hand, as shown in Figure 5(A), in the composite acceleration waveform when specific noise is present, the wave corresponding to the specific noise is superimposed on the wave with a period corresponding to the walking or running pitch, and the wave with a period corresponding to the walking or running pitch may not be clearly visible in this composite acceleration waveform. Therefore, when specific noise is present, it may be difficult to estimate the number of steps from the composite acceleration data.

[0040] As shown in Figure 5(B), when a specific type of noise is present, the U-axis acceleration waveform shows waves with periods corresponding to the arm swing pitch. The number of these waves corresponds to the number of arm swings. Therefore, when a specific type of noise is present, the number of arm swings can be accurately estimated from the U-axis acceleration.

[0041] Therefore, when specific noise is present, it may be preferable to estimate the number of arm swings from U-axis acceleration data and then estimate the number of steps from the estimated number of arm swings, rather than estimating the number of steps from the composite acceleration data. It should be noted that users typically take two steps with one arm swing. Therefore, the estimated number of steps is twice the estimated number of arm swings.

[0042] Here, when a user is running, the amplitude of the wave with a period corresponding to the running pitch in the composite acceleration waveform is considered to be large. Therefore, even if a wave corresponding to a specific noise is superimposed on the wave with a period corresponding to the running pitch in the composite acceleration waveform, the wave with a period corresponding to the running pitch is considered to appear clearly in the composite acceleration waveform. In other words, when a user is running, it is considered possible to estimate the number of steps from the composite acceleration waveform data.

[0043] Furthermore, when the user is neither walking nor running, estimating the number of steps from the number of arm swings estimated from U-axis acceleration data may lead to inaccurate step counts due to arm swings. Therefore, when the user is not walking or when no specific noise is present, it is preferable to estimate the number of steps from composite acceleration data. On the other hand, when the user is walking and specific noise is present, it is preferable to estimate the number of steps from the number of arm swings estimated from U-axis acceleration data.

[0044] The method for determining whether or not a user is walking can be adjusted as appropriate. For example, the walking determination unit 105 determines that a user is walking during the estimated period if the average value of the composite acceleration during the average value calculation period is within the average value reference range, and the difference between the maximum value of the composite acceleration during the difference value calculation period and the minimum value of the composite acceleration during the difference value calculation period is within the difference value reference range.

[0045] The average calculation period is the period for which the average value is calculated, and includes the estimation period. The difference calculation period is the period for which the difference value is calculated, and includes the estimation period. In order to improve the accuracy of discrimination, it is preferable that the average calculation period and the difference calculation period be of a certain length. Therefore, in this embodiment, the average calculation period and the difference calculation period are set to be longer than the estimation period. In this embodiment, the difference calculation period and the average calculation period are the same period. In Figures 4 and 5, T11 shows the estimation period, and T12 shows the average calculation period and the difference calculation period. In this embodiment, the length of the average calculation period and the difference calculation period is the most recent 2 seconds.

[0046] In Figure 4(A), a11 represents the maximum value of the composite acceleration during the difference calculation period, and a12 represents the minimum value of the composite acceleration during the difference calculation period. In this case, the difference value mentioned above is the difference between a11 and a12. In Figure 5(A), a21 represents the maximum value of the composite acceleration during the difference calculation period, and a22 represents the minimum value of the composite acceleration during the difference calculation period. In this case, the difference value mentioned above is the difference between a21 and a22.

[0047] The average value of the combined acceleration when a user is walking is considered to be smaller than the average value of the combined acceleration when a user is running, and larger than the average value of the combined acceleration when a user is stationary. Also, the difference value when a user is walking is considered to be smaller than the difference value when a user is running, and larger than the difference value when a user is stationary. Therefore, if the average value of the combined acceleration is moderate and the difference value of the combined acceleration is moderate, it is determined that the user is walking. The average value reference range is the range in which the average value of the combined acceleration can take when a user is walking. The difference value reference range is the range in which the difference value of the combined acceleration can take when a user is walking.

[0048] The method for determining whether or not specific noise is occurring can be adjusted as appropriate. For example, the noise generation determination unit 106 determines that specific noise is occurring during the estimation target period if the number of maximum points of the composite acceleration during the count calculation period is equal to or greater than the count threshold. The count calculation period is the period for which the number of maximum points of the composite acceleration is calculated, and includes the estimation target period. In order to improve the determination accuracy, it is preferable that the count calculation period be a reasonably long time. Therefore, in this embodiment, the count calculation period is set to be longer than the estimation target period. In this embodiment, the difference value calculation period, the average value calculation period, and the count calculation period are the same period.

[0049] When no specific noise is present, the number of maximum points in the composite acceleration during the counting period corresponds to the number of steps taken during that period. Furthermore, since there is a limit to the pitch during walking, the upper limit of the number of steps taken during the counting period can be determined. Therefore, if the number of maximum points in the composite acceleration during the counting period is greater than or equal to the count threshold, it is determined that specific noise is present during the estimation period. The count threshold is slightly greater than the upper limit of the number of maximum values ​​when the user is walking. For example, if the upper limit of the number of steps per second when the user is walking is less than 3 steps, the count threshold is set to 6.

[0050] In the example shown in Figure 4(A), three local maxima are observed during the counting period T12. Therefore, if composite acceleration data corresponding to the composite acceleration waveform shown in Figure 4(A) is obtained, it is determined that no specific noise is present. On the other hand, in the example shown in Figure 5(A), nine or ten local maxima are observed during the counting period T12. Therefore, if composite acceleration data corresponding to the composite acceleration waveform shown in Figure 5(A) is obtained, it is determined that specific noise is present.

[0051] Incidentally, noise generated by factors other than arm swing may cause local maximums exceeding the count threshold to be observed during the count calculation period. Even in this case, estimating the number of steps from the composite acceleration data is considered to result in low estimation accuracy. Therefore, in this case, the number of steps is estimated from the number of arm swings estimated from the Z-axis acceleration data. In other words, specific noise is not necessarily limited to noise caused by the user's arm swing, but may also be noise caused by factors other than arm swing.

[0052] The method by which the step count estimation unit 103 estimates the number of steps can be adjusted as appropriate. For example, the step count estimation unit 103 estimates the number of steps as the number of first partial waveforms that have a time width within a first reference range and an amplitude greater than or equal to a first amplitude threshold, among the first partial waveforms included in the composite acceleration waveform corresponding to the composite acceleration data after shift processing. The first partial waveform is a waveform that corresponds to one peak in the composite acceleration waveform, and is a waveform from the first timing to the second timing of the composite acceleration waveform.

[0053] The first timing is when the combined acceleration exceeds the average value of the combined accelerations. The first timing is when the combined acceleration waveform crosses zero in the positive direction. The second timing is when the combined acceleration falls below the average value of the combined accelerations. The second timing is when the combined acceleration waveform crosses zero in the negative direction.

[0054] Figure 6 shows a portion of the composite acceleration waveform corresponding to the composite acceleration data after shift processing. In Figure 6, the timing at which the composite acceleration crosses zero while increasing is t31, and the timing at which the composite acceleration crosses zero while decreasing is t32. Therefore, it is ideal to designate t31 as the first timing and t32 as the second timing. However, specifying t31 and t32 involves a large processing load, for example, due to interpolation.

[0055] Therefore, in this embodiment, t33, which is the timing of the first point corresponding to the composite acceleration that changed from a negative value to a positive value, is considered the first timing. Also, t34, which is the timing of the second point corresponding to the composite acceleration that changed from a positive value to a negative value, is considered the second timing. Note that the higher the sampling frequency of the composite acceleration data, the smaller the difference between t31 and t33, and the smaller the difference between t32 and t34. In Figure 6, the first point is shown as a white circle, and the second point is shown as a white square.

[0056] If the first partial waveform is a waveform caused by walking or running, the time width of the first partial waveform corresponds to the walking or running pitch, and the amplitude of the first partial waveform corresponds to the magnitude of the acceleration change caused by walking or running. The range of pitch during walking or running is limited to a certain extent. Also, the acceleration change during walking or running is considered to be somewhat large.

[0057] Therefore, if the time width of the first partial waveform is within the first reference range and the amplitude of the first partial waveform is greater than or equal to the first amplitude threshold, the first partial waveform is considered to be a waveform caused by walking or running, and the number of steps equivalent to one step is estimated for the first partial waveform. The first reference range is defined by the first lower threshold and the first upper threshold. On the other hand, if the time width of the first partial waveform is less than the first lower threshold, the first partial waveform is considered to be a waveform caused by noise. Furthermore, if the time width of the first partial waveform exceeds the first upper threshold, the first partial waveform is considered to be a waveform caused by other factors. Furthermore, if the amplitude of the first partial waveform is less than the first amplitude threshold, the first partial waveform is considered to be a waveform caused by noise.

[0058] In Figure 6, the length of T31, which is the period from t33 to t34, corresponds to the time width of the first partial waveform. Also in Figure 6, the magnitude of a31, which is the maximum value of the composite acceleration in the first partial waveform, corresponds to the amplitude of the first partial waveform. If the length of T31 is greater than or equal to the first lower threshold and less than or equal to the first upper threshold, and the magnitude of a31 is greater than or equal to the first amplitude threshold, Ath, then the first partial waveform is considered to be a waveform caused by walking or running. Note that in Figure 6, the two points indicated by white triangles are points with a composite acceleration of Ath or greater.

[0059] The first lower threshold, first upper threshold, first amplitude threshold, etc., may be predetermined values, or they may be values ​​set according to the user's state estimated from the composite acceleration data. For example, the pitch during walking is considered to be longer than the pitch during running. For this reason, it is preferable to set the first lower threshold during walking to a larger value than the first lower threshold during running. In this case, it can be expected that the misidentification of a first partial waveform caused by noise as a first partial waveform caused by walking will be suppressed, and the misidentification of a first partial waveform caused by running as a first partial waveform caused by noise will be suppressed.

[0060] Furthermore, the magnitude of acceleration changes during walking is considered to be smaller than that during running. For this reason, it is preferable to set the first amplitude threshold during walking to a value smaller than the first amplitude threshold during running. In this case, it is expected that the first partial waveform caused by walking will be suppressed from being mistaken for a first partial waveform caused by noise, and the first partial waveform caused by noise will be suppressed from being mistaken for a first partial waveform caused by running.

[0061] The method for estimating the user's state from the composite acceleration data can be adjusted as appropriate. For example, the pitch during walking is considered to be longer than the pitch during running. Therefore, the number of first timings per unit time identified from the composite acceleration data acquired during walking is less than the number of first timings per unit time identified from the composite acceleration data acquired during running. Thus, the user's state may be considered to be walking when the number of first timings per unit time is less than or equal to a threshold, and the user's state may be considered to be running when the number of first timings per unit time exceeds the threshold.

[0062] Furthermore, the amplitude during walking is considered to be smaller than the amplitude during running. Therefore, the root mean square of the composite acceleration identified from the composite acceleration data acquired during walking is smaller than the root mean square of the composite acceleration identified from the composite acceleration data acquired during running. Thus, the user's state may be considered to be walking when the root mean square of the composite acceleration is less than or equal to the square root threshold, and the user's state may be considered to be running when the root mean square of the composite acceleration exceeds the square root threshold. The user's state may also be determined based on both the number of first timings per unit time and the root mean square of the composite acceleration.

[0063] Furthermore, the method by which the arm swing count estimation unit 104 estimates the number of arm swings is basically the same as the method by which the step count estimation unit 103 estimates the number of steps. For example, the arm swing count estimation unit 104 can estimate the number of arm swings as the number of second partial waveforms among the second partial waveforms included in the specific axis acceleration waveform corresponding to the specific axis acceleration data after shift processing, which have a time width within the second reference range and an amplitude greater than or equal to the second amplitude threshold. The second partial waveform is a waveform that corresponds to one peak in the specific axis acceleration waveform, and is a waveform from the third timing to the fourth timing of the specific axis acceleration waveform.

[0064] The third timing is when the acceleration of a specific axis exceeds the average value of the accelerations of that axis. The third timing is when the specific axis acceleration waveform crosses zero in the positive direction. The fourth timing is when the acceleration of a specific axis falls below the average value of the accelerations of that axis. The fourth timing is when the specific axis acceleration waveform crosses zero in the negative direction.

[0065] If the second partial waveform is a waveform caused by arm swing, the time width of the second partial waveform corresponds to the pitch of the arm swing, and the amplitude of the second partial waveform corresponds to the magnitude of the acceleration change caused by the arm swing. The range of pitch in arm swing is limited to a certain extent. Also, the acceleration change in arm swing is considered to be somewhat large.

[0066] Therefore, if the time width of the second partial waveform is within the second reference range and the amplitude of the second partial waveform is greater than or equal to the second amplitude threshold, the second partial waveform is considered to be a waveform caused by arm swing, and one arm swing is estimated for the second partial waveform. The second reference range is defined by the second lower threshold and the second upper threshold. On the other hand, if the time width of the second partial waveform is less than the second lower threshold, the second partial waveform is considered to be a waveform caused by noise. Also, if the time width of the second partial waveform exceeds the second upper threshold, the second partial waveform is considered to be a waveform caused by other factors. Also, if the amplitude of the second partial waveform is less than the second amplitude threshold, the second partial waveform is considered to be a waveform caused by noise. In this embodiment, the second lower threshold, the second upper threshold, the second amplitude threshold, etc., are predetermined values.

[0067] Next, the step counting process performed by the measuring device 100 will be described with reference to the flowchart in Figure 7. The step counting process is performed, for example, in response to the measuring device 100 receiving a command from the user to start the step counting process.

[0068] First, the control unit 11 of the measuring device 100 acquires 3-axis acceleration data for a specified time (step S101). For example, the control unit 11 acquires newly acquired 1 second of 3-axis acceleration data from the 3-axis acceleration sensor 161 of the detection unit 16. After completing the process in step S101, the control unit 11 generates composite acceleration data (step S102). In other words, the control unit 11 generates new 1 second of composite acceleration data by performing a synthesis process on the newly acquired 1 second of 3-axis acceleration data.

[0069] After completing the processing in step S102, the control unit 11 extracts specific axis acceleration data (step S103). For example, the control unit 11 extracts new 1 second of Z-axis acceleration data from the newly acquired 1 second of 3-axis acceleration data. After completing the processing in step S103, the control unit 11 performs filtering (step S104). For example, the control unit 11 applies a low-pass filter to the newly acquired 1 second of composite acceleration data to obtain 1 second of composite acceleration data from which high-frequency components have been removed. The control unit 11 also applies a low-pass filter to the newly acquired 1 second of specific axis acceleration data to obtain 1 second of specific axis acceleration data from which high-frequency components have been removed.

[0070] After completing the processing in step S104, the control unit 11 performs an offset process (step S105). For example, the control unit 11 performs an offset process on the composite acceleration data for one second from which the high-frequency components have been removed, and obtains composite acceleration data for one second from which the DC components have been removed. The control unit 11 also performs an offset process on the specific axis acceleration data for one second from which the high-frequency components have been removed, and obtains specific axis acceleration data for one second from which the DC components have been removed.

[0071] After completing the process in step S105, the control unit 11 executes the step count estimation process (step S106). The step count estimation process will be described in detail below with reference to Figure 8.

[0072] First, the control unit 11 identifies the first timing (step S201). For example, the control unit 11 identifies all points in the composite acceleration waveform shown by the composite acceleration data for one second from which the DC component has been removed, where the value changes from negative to positive. After completing the process in step S201, the control unit 11 identifies the second timing (step S202). For example, the control unit 11 identifies all points in the composite acceleration waveform shown by the composite acceleration data for one second from which the DC component has been removed, where the value changes from positive to negative.

[0073] After completing the processing in step S202, the control unit 11 determines a first reference range (step S203). For example, the control unit 11 estimates the user's state from the composite acceleration data for 1 second from which the DC component has been removed, and determines the first reference range based on the estimated user's state. After completing the processing in step S203, the control unit 11 determines a first amplitude threshold (step S204). For example, the control unit 11 determines the first amplitude threshold based on the estimated user's state.

[0074] After completing the processing in step S204, the control unit 11 determines whether or not there are any unselected first partial waveforms (step S205). In other words, the control unit 11 determines whether or not there are any first partial waveforms included in the newly acquired 1-second composite acceleration waveform that were not selected in step S206. Selecting a first partial waveform corresponds to selecting a pair of a first timing and a second timing that follows this first timing.

[0075] Here, the first partial waveform included in the newly acquired 1-second composite acceleration waveform is the same as the first partial waveform included in the newly acquired 1-second composite acceleration waveform for the second timing. In other words, the first partial waveform included in the newly acquired 1-second composite acceleration waveform is not limited to the first partial waveforms included in the newly acquired 1-second composite acceleration waveform for both the first and second timings, but also includes the first partial waveform included in the previously acquired 1-second composite acceleration waveform for the first timing and the first partial waveform included in the newly acquired 1-second composite acceleration waveform for the second timing.

[0076] If the control unit 11 determines that there are no unselected first partial waveforms (step S205: NO), it completes the step count estimation process. If the control unit 11 determines that there are unselected first partial waveforms (step S205: YES), it selects a first partial waveform (step S206). After completing the process in step S206, the control unit 11 determines whether the time width of the first partial waveform is within the first reference range (step S207). If the control unit 11 determines that the time width of the first partial waveform is within the first reference range (step S207: YES), it determines whether the amplitude of the first partial waveform is greater than or equal to the first amplitude threshold (step S208).

[0077] If the control unit 11 determines that the amplitude of the first partial waveform is greater than or equal to the first amplitude threshold (step S208: YES), it counts one step for the first partial waveform (step S209). If the control unit 11 determines that the time width of the first partial waveform is not within the first reference range (step S207: NO), if it determines that the amplitude of the first partial waveform is not greater than or equal to the first amplitude threshold (step S208: NO), or if it has completed the process in step S209, it returns to step S205.

[0078] After completing the step count estimation process in step S106, the control unit 11 executes the arm swing count estimation process. The arm swing count estimation process will be described in detail below with reference to Figure 9. In the arm swing count estimation process, newly acquired specific axis acceleration data for a specified time period is processed in chronological order of acquisition time. This specific axis acceleration data is the specific axis acceleration data after offset processing.

[0079] First, the control unit 11 determines whether the positive zero-crossing flag is set (step S301). The positive zero-crossing flag is a flag that is set when a positive zero-crossing is detected in the specific-axis acceleration waveform indicated by the specific-axis acceleration data. The detection of a positive zero-crossing corresponds to the detection of the third timing. The positive zero-crossing flag is set in the arm swing count estimation process for the specific-axis acceleration data for the specified time obtained last time, or in the arm swing count estimation process for the newly obtained specific-axis acceleration data for the specified time.

[0080] When the control unit 11 determines that the positive zero-crossing flag is not set (step S301: NO), it determines whether Acc(t - 1) ≤ 0 < Acc(t) (step S302). Acc(t) indicates the data being processed among the specific-axis acceleration data. Acc(t - 1) indicates the data obtained one before the data being processed among the specific-axis acceleration data. That is, Acc(t - 1) ≤ 0 < Acc(t) corresponds to the detection of a positive zero-crossing in the data being processed.

[0081] When the control unit 11 determines that Acc(t - 1) ≤ 0 < Acc(t) (step S302: YES), it sets the positive zero-crossing flag (step S303). When the control unit 11 determines that the positive zero-crossing flag is set (step S301: YES), it increments the provisional time width (step S304). The provisional time width is a provisional value of the time width of the second partial waveform. The provisional time width corresponds to the time from the time when the data with the positive zero-crossing flag set was obtained to the time when the data being processed is obtained. Incrementing the provisional time width corresponds to increasing the provisional time width by one sampling period.

[0082] When the control unit 11 finishes the process of step S304, it determines whether Acc(t) ≥ the second amplitude threshold value (step S305). That is, the control unit 11 determines whether the currently processed data is greater than or equal to the second amplitude threshold value. When the control unit 11 determines that Acc(t) ≥ the second amplitude threshold value (step S305: YES), it sets the intensity flag (step S306). The intensity flag is a flag that is set when the specific-axis acceleration data indicating the second partial waveform currently being processed contains data greater than or equal to the second amplitude threshold value.

[0083] When the control unit 11 determines that Acc(t) < the second amplitude threshold value (step S305: NO), or when it finishes the process of step S306, it determines whether Acc(t) ≤ 0 < Acc(t - 1) (step S307). Acc(t) ≤ 0 < Acc(t - 1) corresponds to the detection of a zero crossing in the negative direction in the currently processed data. When a zero crossing in the negative direction is detected in the currently processed data, the second partial waveform currently being processed is completely specified.

[0084] When the control unit 11 determines that Acc(t) ≤ 0 < Acc(t - 1) (step S307: YES), it determines whether the tentative time width ≥ the second lower threshold value (step S308). The tentative time width ≥ the second lower threshold value corresponds to the fact that the second partial waveform currently being processed is wide enough not to be assumed as noise. When the control unit 11 determines that the tentative time width ≥ the second lower threshold value (step S308: YES), it determines whether the intensity flag is set (step S309).

[0085] When the control unit 11 determines that the intensity flag is set (step S309: YES), it increments the tentative number of arm swings (step S310). The tentative number of arm swings is a tentative value of the number of arm swings estimated from the specific-axis acceleration data for a specified time newly acquired in the current arm swing number estimation process. Incrementing the tentative number of arm swings corresponds to increasing the tentative number of arm swings by one. The tentative number of arm swings is initialized to 0 when the arm swing number estimation process starts.

[0086] When the control unit 11 determines that Acc(t) ≤ 0 < Acc(t−1) is not satisfied (step S307: NO), it determines whether the provisional time width > the second upper threshold value (step S311). The fact that the provisional time width > the second upper threshold value corresponds to the fact that the current second partial waveform is wide enough not to be assumed as a waveform caused by arm swinging.

[0087] When the control unit 11 determines that the provisional time width ≥ the second lower threshold value is not satisfied (step S308: NO), when it determines that the intensity flag is not set (step S309: NO), when the process of step S310 is completed, or when it determines that the provisional time width > the second upper threshold value (step S311: YES), it initializes each flag and the provisional time width (step S312). That is, the control unit 11 resets the positive direction zero-crossing flag and the intensity flag, and resets the provisional time width to 0.

[0088] When the control unit 11 determines that Acc(t−1) ≤ 0 < Acc(t) is not satisfied (step S302: NO), when it determines that the provisional time width > the second upper threshold value is not satisfied (step S311: NO), or when the process of step S303 or step S312 is completed, it determines whether all the data for the specified time have been processed (step S313). That is, the control unit 11 determines whether the processing of all the specific-axis acceleration data for the newly acquired specified time has been completed.

[0089] When the control unit 11 determines that not all the data for the specified time have been processed (step S313: NO), it proceeds to the processing of the next data (step S314). That is, the control unit 11 increases t that designates the data to be processed by 1. When the control unit 11 completes the process of step S314, it returns the process to step S301. When the control unit 11 determines that all the data for the specified time have been processed (step S313: YES), it completes the arm swing count estimation process.

[0090] When the control unit 11 completes the arm swing count estimation process in step S107, it determines whether or not the user is walking (step S108). For example, if the average value of the composite acceleration over the most recent 2 seconds is within the average value reference range, and the difference between the maximum value and the minimum value of the composite acceleration over the most recent 2 seconds is within the difference value reference range, the control unit 11 determines that the user was walking in the most recent 1 second.

[0091] When the control unit 11 determines that the user is walking (step S108: YES), it determines whether or not a specific noise is occurring (step S109). For example, if the control unit 11 determines that a specific noise has occurred in the last second, it determines that a specific noise has occurred in the last second if the number of maximum points of the combined acceleration in the last two seconds is six or more.

[0092] If the control unit 11 determines that the user is not walking (step S108: NO), or if it determines that no specific noise is occurring (step S109: NO), it adds the estimated number of steps to the cumulative number of steps (step S110). For example, the control unit 11 adds the estimated number of steps for the most recent second, which was estimated in the step count estimation process, to the cumulative number of steps.

[0093] When the control unit 11 determines that a specific noise is occurring (step S109: YES), it adds twice the estimated number of arm swings to the cumulative number of steps (step S111). For example, the control unit 11 adds twice the estimated number of arm swings in the most recent second, which was estimated in the arm swing estimation process, to the cumulative number of steps. When the control unit 11 completes the processing in step S110 or step S111, it returns to step S101.

[0094] Next, with reference to Figure 10, the error in step count measurement by the measuring device 100 will be explained. Figure 10 shows the measurement error when step count measurements were performed for various trial patterns using the measuring device 100 and the measuring device related to the comparative example. The step count measurement error is expressed as a percentage (MT) / T, where M is the number of steps obtained by step count measurement and T is the true number of steps. The trial patterns are defined by combinations of the type of walking or running, the location of walking or running, and the direction of movement in walking or running that involves vertical movement. Figure 10 shows the measurement results when step counts were measured using 28 subjects for 11 trial patterns.

[0095] In Figure 10, the average measurement error of the measuring device related to the comparative example is shown by a white rectangle, and the average measurement error of measuring device 100 is shown by a black rectangle. Lines extending horizontally through the white or black rectangles indicate the range of average errors for each subject. Measuring device 100 performs step count estimation processing using composite acceleration data and arm swing count estimation processing using specific axis acceleration data, while the measuring device related to the comparative example performs step count estimation processing using composite acceleration data only. In other words, measuring device 100 measures the number of steps based on the estimation result that is estimated to have higher estimation accuracy between the step count estimation result and the arm swing count estimation result. On the other hand, the measuring device related to the comparative example measures the number of steps based on the step count estimation result.

[0096] As shown in Figure 10, overall, the measurement error with measuring device 100 was smaller than the measurement error with the measuring device of the comparative example. For example, the overall measurement error with measuring device 100 was 0.16 ± 4.00%, while the overall measurement error with the measuring device of the comparative example was 0.31 ± 4.44%. Furthermore, in many trial patterns, it was confirmed that the measurement error with measuring device 100 showed less variability among subjects than the measurement error with the measuring device of the comparative example.

[0097] Furthermore, some trial patterns were observed in which the measurement error of the measuring device 100 was greater than that of the measuring device in the comparative example. For example, in the case of descending stairs, the measurement error of the measuring device 100 may be slightly greater than that of the measuring device in the comparative example. However, in the main trial patterns expected when measuring steps, the measurement error of the measuring device 100 was found to be smaller than that of the measuring device in the comparative example. The main trial patterns include walking (on flat ground), walking (uphill), walking (downhill), running (on flat ground), running (uphill), and running (downhill). Therefore, it can be concluded that, in practice, the measurement error of the measuring device 100 is smaller than that of the measuring device in the comparative example.

[0098] In this embodiment, if the user is not walking or if no specific noise is occurring, the number of steps estimated based on the composite acceleration data is used. On the other hand, if the user is walking and specific noise is occurring, twice the number of arm swings estimated based on the specific axis acceleration data is used as the user's step count. Therefore, according to this embodiment, the accuracy of step count measurement can be improved.

[0099] Furthermore, in this embodiment, if the average value of the composite acceleration is within the average value reference range, and the difference between the maximum value and the minimum value of the composite acceleration is within the difference value reference range, it is determined that the user is walking. Therefore, according to this embodiment, it is possible to determine with high accuracy whether or not the user is walking, and the accuracy of step counting can be further improved.

[0100] Furthermore, in this embodiment, if the number of maximum points of composite acceleration during the count calculation period, which includes the estimated period, is greater than or equal to the count threshold, it is determined that specific noise has occurred during the estimated period. Therefore, according to this embodiment, it is possible to determine with high accuracy whether or not specific noise has occurred, and the accuracy of step count measurement can be further improved.

[0101] Furthermore, in this embodiment, if it is determined that the time width of the first partial waveform in the composite acceleration waveform is within the first reference range and the amplitude of the first partial waveform is greater than or equal to the first amplitude threshold, the number of steps for one step is estimated for the first partial waveform. Therefore, according to this embodiment, the number of steps can be estimated with high accuracy based on the composite acceleration data, and the accuracy of step count measurement can be further improved.

[0102] Furthermore, in this embodiment, if the time width of the second partial waveform in the specific axis acceleration waveform is determined to be within the second reference range and the amplitude of the second partial waveform is greater than or equal to the second amplitude threshold, one arm swing is estimated for the second partial waveform. Therefore, according to this embodiment, the number of arm swings can be estimated with high accuracy based on the specific axis acceleration data, further improving the accuracy of step count measurement.

[0103] (modified version) Although embodiments have been described above, various modifications and applications are possible. It is arbitrary which parts of the configuration, function, and operation described in the above embodiments are adopted. Furthermore, additional configurations, functions, and operations may be adopted in addition to those described above. Also, the configurations, functions, and operations described in the above embodiments can be freely combined.

[0104] The step count estimation process is not limited to the examples described in the embodiments. For example, the embodiments described an example in which the first reference range, the first amplitude threshold, etc., are determined according to the user's state estimated from the composite acceleration data. The first reference range, the first amplitude threshold, etc., may be predetermined values. Furthermore, the embodiments described an example in which it is determined whether or not to estimate the number of steps for one step for the first partial waveform based on both the time width and the amplitude of the first partial waveform. It is also possible to determine whether or not to estimate the number of steps for one step for the first partial waveform based on either the time width or the amplitude of the first partial waveform. Thus, various processes for estimating the number of steps based on composite acceleration data can be employed as the step count estimation process.

[0105] Similarly, the arm swing count estimation process is not limited to the examples described in the embodiments. For example, the embodiments described an example in which an offset process is performed on specific axis acceleration data. Offset processing or the like is not required to be performed on specific axis acceleration data. Also, the embodiments described an example in which it is determined whether or not to estimate one arm swing for the second partial waveform based on both the time width and the amplitude of the second partial waveform. It is also possible to determine whether or not to estimate one arm swing for the second partial waveform based on either the time width or the amplitude of the second partial waveform. Thus, various processes can be employed to estimate the number of arm swings based on specific axis acceleration data as the arm swing count estimation process.

[0106] In this embodiment, an example was described in which the estimation period is the most recent 1 second, and the average calculation period, difference calculation period, and count calculation period are all the most recent 2 seconds. However, the estimation period, average calculation period, difference calculation period, and count calculation period are not limited to these periods. For example, the estimation period may be less than the most recent 1 second, or it may be more than the most recent 1 second. Also, the average calculation period, difference calculation period, and count calculation period may be less than the most recent 2 seconds, or they may be more than the most recent 2 seconds. Furthermore, the average calculation period, difference calculation period, and count calculation period do not have to encompass the entire estimation period. The average calculation period, difference calculation period, and count calculation period may be different periods from each other.

[0107] In the above embodiment, the control unit 11 functioned as the respective components shown in Figure 3 by the CPU executing a program stored in the ROM or storage unit 12. However, in this disclosure, the control unit 11 may be dedicated hardware. Dedicated hardware includes, for example, a single circuit, a composite circuit, a programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof. If the control unit 11 is dedicated hardware, each function of the respective components may be implemented by separate hardware, or the functions of each component may be implemented together by a single piece of hardware. Furthermore, some of the functions of each component may be implemented by dedicated hardware, while other parts may be implemented by software or firmware. In this way, the control unit 11 can implement the above-mentioned functions by hardware, software, firmware, or a combination thereof.

[0108] It is also possible to make an existing personal computer or information terminal or other computer function as the measuring device 100 by applying an operating program that defines the operation of the measuring device 100 related to this disclosure to that computer. Furthermore, the method of distributing such a program is arbitrary; for example, it may be distributed by storing it on a computer-readable recording medium such as a CD-ROM (Compact Disk ROM), DVD (Digital Versatile Disk), MO (Magneto Optical Disk), or memory card, or it may be distributed via a communication network such as the Internet.

[0109] This disclosure allows for various embodiments and modifications without departing from the broad spirit and scope of this disclosure. Furthermore, the embodiments described above are for illustrative purposes only and do not limit the scope of this disclosure. In other words, the scope of this disclosure is indicated by the claims, not by the embodiments. Various modifications made within the scope of the claims and the equivalent significance of the disclosure are considered to be within the scope of this disclosure. [Explanation of Symbols]

[0110] 11 Control unit, 12 Storage unit, 13 Display unit, 14 Operation reception unit, 15 Communication unit, 16 Detection unit, 100 Measurement device, 101 Data acquisition unit, 102 Data processing unit, 103 Step count estimation unit, 104 Arm swing count estimation unit, 105 Walking detection unit, 106 Noise generation detection unit, 107 Step count measurement unit, 108 Skin control unit, 161 3-axis acceleration sensor

Claims

1. A data acquisition means for acquiring 3-axis acceleration data for 3 axes output by a 3-axis acceleration sensor attached to the user, A step count estimation means that estimates the user's step count based on composite acceleration data obtained from the three-axis acceleration data, An arm swing count estimation means estimates the number of times the user swings their arm based on specific axis acceleration data from the three-axis acceleration data that corresponds to the direction in which the user's arm is swung, The system includes a step counting means that measures the user's steps by switching between the user's step count estimated by the step count estimation means and the user's arm swing count estimated by the arm swing count estimation means, according to predetermined criteria. Measuring device.

2. A means for determining whether the user is walking or not, A noise generation detection means for determining whether or not a specific noise is occurring, Equipped with, If the step counting means determines that the user is not walking, or that the specific noise is not occurring, it measures the user's steps by adopting the steps estimated by the step counting means. If the user is walking and the specific noise is occurring, it measures the user's steps by adopting twice the number of arm swings estimated by the arm swing counting means. The measuring device according to claim 1.

3. The walking determination means determines that the user was walking during the estimated period if the average value of the composite acceleration during the average value calculation period including the estimated period is within the average value reference range, and the difference between the maximum value of the composite acceleration during the difference value calculation period including the estimated period and the minimum value of the composite acceleration during the difference value calculation period is within the difference value reference range. The measuring device according to claim 2.

4. The noise generation determination means determines that the specific noise occurred during the estimation period if the number of maximum points of the composite acceleration during the count calculation period, which includes the estimation period, is equal to or greater than the count threshold. The measuring device according to claim 2 or 3.

5. The step count estimation means determines that the time width of the first partial waveform, which is the waveform in the composite acceleration waveform corresponding to the composite acceleration data from the first timing, when the composite acceleration exceeds the average value of the composite acceleration, to the second timing, when the composite acceleration falls below the average value of the composite acceleration, is within a first reference range, and that the amplitude of the first partial waveform is greater than or equal to a first amplitude threshold, and estimates the number of steps for one step with respect to the first partial waveform. A measuring device according to any one of claims 1 to 3.

6. The arm swing count estimation means determines that the time width of the second partial waveform, which is the waveform from the third timing (the timing when the acceleration of a specific axis, which is the acceleration of a specific axis) to the fourth timing (the timing when the acceleration of a specific axis falls below the average value of the specific axis acceleration) in the specific axis acceleration waveform corresponding to the specific axis acceleration data, is within the second reference range, and that the amplitude of the second partial waveform is greater than or equal to the second amplitude threshold, then estimates one arm swing for the second partial waveform. A measuring device according to any one of claims 1 to 3.

7. The 3-axis acceleration data for all three axes output by the 3-axis acceleration sensor attached to the user is acquired. Based on the composite acceleration data obtained from the three-axis acceleration data, the user's step count is estimated. Based on the specific axis acceleration data corresponding to the direction in which the user's arm is swung, among the three-axis acceleration data, the number of times the user's arm swings is estimated. The system measures the user's steps by switching between the estimated number of steps taken by the user and the estimated number of arm swings by the user, according to predetermined criteria. Measurement method.

8. Computers, A data acquisition means for acquiring 3-axis acceleration data for 3 axes output by a 3-axis acceleration sensor attached to the user. Step count estimation means that estimates the user's step count based on composite acceleration data obtained from the three-axis acceleration data, An arm swing count estimation means that estimates the number of times the user swings their arm based on specific axis acceleration data from the three-axis acceleration data that corresponds to the direction in which the user's arm is swung. The step counting means functions by switching between and using the user's step count estimated by the step count estimation means and the user's arm swing count estimated by the arm swing count estimation means, according to predetermined criteria. program.

Images