JP5699717B2 - Step detection device, electronic device and program - Google Patents

Description

  The present invention relates to a step detection device, an electronic device, a program, and the like.

  Pedometers using acceleration sensors are widely used. The pedometer is intended to measure the number of steps accompanying the user's walking, but as long as the acceleration sensor is used, a change in acceleration caused by factors other than walking may be detected as the number of steps. For example, even when moving with a vehicle such as a car while wearing a pedometer, the acceleration sensor detects a change in acceleration information due to acceleration, deceleration, and other factors, and counts it as the number of steps.

  Patent Document 1 discloses a technique for suppressing erroneous counting due to vibration of a vibrator or the like. In Patent Document 1, the erroneous count is suppressed by performing Fourier transform and analyzing the frequency. Patent Document 2 discloses a technique for distinguishing between counting based on walking and counting based on other factors based on the difference between a detected signal value and a past detected peak level. Patent Document 3 discloses a technique for determining information about energy by integrating acceleration and determining detection of the number of steps by comparing the energy information with a predetermined threshold.

JP 2010-071777 A JP 2009-064136 A JP 2008-171347 A

  In the method of Patent Document 1, since it is necessary to perform Fourier transform, there is a problem that the processing load is heavy. In the method of Patent Document 2, it is difficult to detect when a pedometer is placed in a pants pocket. This is because, when the pedometer is put in a pants pocket, the variation in the detection peak becomes large. Conversely, it is conceivable that the variation in detection peaks is smaller when moving in an automobile or the like, and in this case, it is difficult to set a threshold value for appropriately separating walking and vehicle movement. Further, in the method of Patent Literature 3, it is difficult to make a determination when a pedometer is put in a pants pocket, as in Patent Literature 2. This is because, for example, when it is worn in a pocket of a trouser, the variation in energy increases, so it is difficult to distinguish between detection by walking and detection by vehicle movement.

  According to some aspects of the present invention, the determination based on the feature amount for determining whether or not the vehicle is moving is performed, and the count is reserved based on the determination result, thereby preventing erroneous counting during vehicle movement. A step detection device, an electronic device, a program, and the like can be provided.

  One aspect of the present invention is a step number detection unit that detects the number of steps based on a sensor signal from an acceleration sensor, a walking time measurement unit that measures a continuous walking time from the start of walking, and the step number detection in the step number detection unit. A step counting unit that counts the number of steps based on the result of the step, a feature amount extracting unit that extracts a feature amount for determining whether or not the vehicle is moving based on the sensor signal, and the feature amount A determination unit that determines whether the user is moving based on the feature amount extracted by the extraction unit, and the step count unit includes a continuous walk time from a given effective step determination period. When it is short, the count of the number of steps detected in the continuous walking time is reset, and the step count counting unit is based on the determination result in the determination unit and the user is moving the vehicle If it is a constant related to the step detection apparatus for extending the effective number of steps determination period.

  In one aspect of the present invention, when it is determined that the vehicle is moving based on the determination using the feature amount, the effective step determination period is extended. Therefore, in view of the fact that the effective step count determination period is related to the determination and resetting of the counter, it is possible to suppress erroneous counting due to vehicle movement by extending the effective step count determination period.

  In one aspect of the present invention, the step count detection unit obtains an evaluation value of the walking based on the sensor signal at a timing specified by a timing at which a peak value is detected, and the determination unit evaluates If the evaluated value determined as walking among the evaluated values in the period is larger than a given threshold, it may be determined that the user has not moved the vehicle.

  Thereby, when the value to be evaluated is large, for example, it is possible to determine that the walking is performed without performing processing using another feature amount.

  In the aspect of the invention, the feature amount extraction unit extracts the number of zero crosses of the twice-integrated signal value of the acceleration power vector as the feature amount, and the determination unit includes the evaluation value of the evaluation value in the evaluation period. If it is determined that the evaluation value determined as walking is smaller than a given threshold, the number of steps detected by the step detection unit is compared with the number of zero crossings, and the comparison result is obtained. Based on this, it may be determined whether the user is moving the vehicle.

  Thereby, when the value to be evaluated is small, it is possible to distinguish between walking and vehicle movement by comparing the number of zero crossings and the number of detected steps.

  In the aspect of the invention, the feature amount extraction unit extracts the number of zero crosses of the twice integrated signal value of the acceleration power vector as the feature amount, and the determination unit detects the number of steps detected by the step number detection unit. And the number of times of zero crossing may be compared to determine whether the user is moving based on the comparison result.

  This makes it possible to distinguish walking and vehicle movement by comparing the number of zero crossings and the number of detected steps.

  In one aspect of the present invention, the step count detection unit obtains an evaluation value of the walking based on the sensor signal at a timing specified by a timing at which a peak value is detected, and the determination unit Based on the relationship between the evaluated value determined as walking among the evaluated values and the evaluated value determined as not walking, it may be determined whether the user is moving the vehicle.

  Accordingly, it is possible to determine whether or not the vehicle is moving based on the evaluated value determined as walking and the evaluated value determined as not walking.

  Moreover, in one aspect of the present invention, the determination unit determines that a value obtained based on a sum of evaluated values determined to be walking among the evaluated values in an evaluation period is not walking. It may be determined whether the user is moving the vehicle by obtaining a ratio with a value obtained based on the sum of the evaluated values and comparing the obtained ratio with a given threshold value.

  As a result, it is possible to use a ratio of two values as an example of the relationship between the evaluated value determined as walking and the evaluated value determined as not walking.

  Further, in one aspect of the present invention, the step count unit includes a temporary counter and a final output counter, and the step count unit, when the continuous walking time is shorter than the effective step number determination period, The number of steps may be counted in a temporary counter, and when the continuous walking time exceeds the effective number of steps determination period, the value of the temporary counter may be added to the value of the final output counter.

  As a result, the counter can be determined and the reset process can be realized by using the temporary counter and the final output counter according to the effective step count determination period.

  In one embodiment of the present invention, the step detection unit sets a threshold based on a step detection prediction period and an elapsed time since the previous step detection, and is specified by a timing at which a peak value is detected. The number of steps may be detected when an evaluated value obtained based on the sensor signal at the timing is larger than the threshold value.

  This makes it possible to set a threshold based on the prediction detection cycle and the elapsed time since the previous step detection.

  According to another aspect of the present invention, there is provided a step count detecting unit that detects a step count based on a sensor signal from an acceleration sensor, and a step count that performs a step count process based on a result of the step count detection by the step count detecting unit. The step detection unit sets a threshold based on the step detection detection cycle and the elapsed time since the previous step detection, and at a timing specified by the timing at which the peak value is detected. The present invention relates to a step detection device that detects the number of steps when an evaluated value obtained based on the sensor signal is larger than the threshold value.

  In the aspect of the invention, the step count detection unit may include a lookup table in which an output value is determined based on the prediction detection cycle and the elapsed time from the previous step count detection. The output value may be set as the threshold value.

  Thereby, the threshold value can be set using the lookup table.

In one aspect of the present invention, the step detection unit detects the first to N-th step detection times t _{1 to} t _N and uses the N-th prediction detection cycle used for the (N + 1) -th step detection process. You may obtain | require _TN as _TN = tN _-1- tN _-2 .

  This makes it possible to appropriately set the prediction detection cycle even in the case where the step count detection device is mounted in the pocket of the pants.

  Another aspect of the invention relates to an electronic apparatus including the step count detection device according to any one of claims 1 to 11.

  According to another aspect of the present invention, a step number detection unit that detects the number of steps based on a sensor signal from the acceleration sensor, a walking time measurement unit that measures a continuous walking time from the start of walking, and the step number detection unit A step counting unit that performs a step counting process based on the result of the step detection, a feature amount extracting unit that extracts a feature amount for determining whether or not the vehicle is moving, based on the sensor signal; Based on the feature amount extracted by the feature amount extraction unit, the computer functions as a determination unit that determines whether the user is moving the vehicle, and the step count counting unit is provided with the continuous walking time given When it is shorter than the effective step count determination period, the count of the number of steps detected in the continuous walking time is reset, and the step count counting unit is based on the determination result in the determination unit If the user is determined to be moving vehicle, there is provided a program to extend the effective number of steps determination period.

  According to another aspect of the present invention, there is provided a step count detecting unit that detects a step count based on a sensor signal from an acceleration sensor, and a step count that performs a step count process based on a result of the step count detection by the step count detecting unit. The computer functions as a unit, and the step detection unit sets a threshold based on the predicted detection cycle of the step count and the elapsed time since the previous step detection, and is identified by the timing at which the peak value is detected This relates to a program for detecting the number of steps when the value to be evaluated obtained based on the sensor signal is larger than the threshold value.

The system configuration example of the step detection device of this embodiment. The figure explaining the change of the acceleration power vector by performing the filter process by a band pass filter. The figure explaining a lower peak value. The figure explaining the example whose 1st to-be-evaluated value calculation process is not appropriate. Lookup table that determines the threshold. The table which determines the threshold value at the time of starting a walk from a stop state. The figure explaining the walk cycle at the time of normal walk. The figure explaining the walk cycle at the time of mounting | wearing the pocket of a trouser with a step count detection apparatus. The figure explaining the acceleration power vector at the time of normal walking, and the relationship of the twice integral signal of an acceleration power vector. The figure explaining the relationship between the acceleration power vector at the time of vehicle movement, and the twice integral signal of an acceleration power vector. The figure explaining the relationship between the acceleration power vector at the time of mounting | wearing the pocket of a trouser with a step count detection apparatus, and the twice integral signal of an acceleration power vector. The figure explaining the state of the false detection of the step count by the vibration at the time of vehicle movement. The flowchart for demonstrating the determination process using a feature-value. The flowchart for demonstrating the detail of step count processing. 1 is a system configuration example of an electronic apparatus according to an embodiment.

  Hereinafter, this embodiment will be described. In addition, this embodiment demonstrated below does not unduly limit the content of this invention described in the claim. In addition, all the configurations described in the present embodiment are not necessarily essential configuration requirements of the present invention.

1. First, the method of this embodiment will be described. Step detection devices (pedometers) using acceleration sensors are widely used, but since these step detection devices detect the number of steps based on acceleration fluctuations, the acceleration caused by factors other than walking (vehicle movement, etc.) Variations may also be counted as steps taken by walking. Therefore, various methods for distinguishing between walking and other factors have been proposed, as described in Patent Documents 1 to 3 described above.

  However, the method of performing Fourier transform has a problem that the processing load of the Fourier transform processing is heavy. Also, the method of comparing the detected peak level with the signal level or comparing with energy causes a problem when a step count detecting device is attached to the pants pocket. This is because when the step count detection device is attached to the pants pocket, the motion state of the step count detection device is greatly different between the stepping with the right foot and the stepping with the left foot, so that the detection peak and the variation in energy become large.

  In view of this, the present applicant proposes a method for distinguishing between walking and vehicle movement by obtaining a feature amount for determining whether or not the vehicle is moving, and making a determination based on the feature amount. In the present embodiment, an evaluation value (evaluation value) is calculated from acceleration information (for example, a power vector described later) from the acceleration sensor, and the number of steps is detected based on the evaluation value. At the same time, a feature amount is obtained from acceleration information or an evaluated value, and it is determined whether the number of steps is due to walking or vehicle movement. When it is determined that the number of steps is detected by moving the vehicle, processing such as reservation or cancellation of the count is performed when the number of steps is counted. By doing in this way, it becomes possible to suppress the erroneous count at the time of vehicle movement. The processing such as the reservation of the count is performed by extending the effective step number determination period, etc., and details of the effective step number determination period will be described later.

  Hereinafter, after describing the system configuration example of the present embodiment, the step count processing in the present embodiment will be described. Specifically, the step count processing will be described by dividing it into pre-processing, step count detection processing, feature amount calculation and determination processing, and count processing. Finally, the operation and effect of this embodiment will be described.

2. System Configuration Example of this Embodiment A configuration example of a pedometer (step count measurement system) including the step count detection apparatus 100 according to this embodiment will be described with reference to FIG. The pedometer includes a sensor unit 10, a preprocessing unit 110, a step count detection device 100, and a step count display unit 20. The step count detection apparatus 100 includes a step count detection unit 120, a feature amount extraction unit 130, a determination unit 140, a walking time measurement unit 150, and a step count counting unit 160. However, the pedometer 100 and the pedometer are not limited to the configuration shown in FIG. 1, and various modifications such as omitting some of these components or adding other components are possible.

  The sensor unit 10 includes a sensor used for walking count processing. Here, the sensor unit 10 includes an acceleration sensor. A sensor other than the acceleration sensor may be included.

  The preprocessing unit 110 acquires sensor information from the sensor unit 10 and performs preprocessing. Specifically, a process for obtaining a power vector from triaxial acceleration information and a filter process using a bandpass filter (hereinafter, referred to as BPF as appropriate) are performed.

  The step count detection unit 120 performs step count detection processing. Here, the number of steps detection process is, for example, a process for detecting a timing at which walking is considered based on a signal value after preprocessing. Therefore, as an example, the output from the step count detection unit 120 to the walking time measurement unit 150 and the step count counting unit 160 is performed in the form of a step count detection pulse.

  The step detection unit 120 includes a peak detection unit 121, an evaluated value calculation unit 123, and a threshold determination unit 125. The peak detection unit 121 detects a peak value from signal values after preprocessing. The to-be-evaluated value calculation unit 123 calculates an to-be-evaluated value used for step count detection from the signal value at the timing when the peak value is detected. The threshold determination unit 125 adaptively sets a threshold, and detects the number of steps by comparing the set threshold and the evaluated value. Specific processing of the peak detection unit 121, the evaluated value calculation unit 123, and the threshold determination unit 125 will be described later.

  The feature amount extraction unit 130 calculates a feature amount for determining whether or not the vehicle is moving. Details of the feature amount and the calculation method will be described later. For example, the signal value after the preprocessing from the preprocessing unit 110, the value to be evaluated from the evaluation value calculation unit 123, or the step count detection pulse from the threshold determination unit 125 It is possible to calculate based on the above.

  The determination unit 140 determines whether or not the number of steps detected during the evaluation period is due to vehicle movement based on the feature amount calculated by the feature amount extraction unit 130. Details of the evaluation period and a specific determination method will be described later.

  The walking time measuring unit 150 measures the continuous walking time. Here, the continuous walking time represents a time during which walking is continuously detected.

  The step count unit 160 counts the number of steps. In the step detection process in the step detection unit 120, it is not particularly determined whether the vehicle is moving. Whether the step count detection result is reflected in the step count displayed on the step count display unit 20 is determined by the step count counting unit 160 in consideration of the determination result in the determination unit 140.

  The step count counting unit 160 includes a temporary counter 161 and a final output counter 163. When the first walk is detected from the stationary state, first, the temporary counter 161 counts. The counting by the temporary counter 161 is performed until the continuous walking time exceeds the effective step number determination period, and when the effective step number determination period is exceeded, the count value of the temporary counter 161 is added to the final output counter 163. The count value of the final output counter 163 is transmitted to the step count display unit 20 and presented to the user.

  In other words, until the continuous walking time exceeds the effective number of steps determination period, even if the number of steps is detected, it is a tentative count, and is determined as a counter presented to the user for the first time after the effective number of steps determination period is exceeded. It will be. In the present embodiment, an erroneous count due to vehicle movement is suppressed by extending the effective number of steps determination period. However, extending the effective number of steps determination period leads to an increase in time until the detection of the number of steps is determined as a count. If the time until confirmation is extended, there is a grace period for acquiring information on whether or not the provisional count is due to vehicle movement, and as a result, count suppression processing such as resetting the temporary counter 161 can be performed. The nature will increase. For example, when the effective step count determination period exceeds the allowable value (or when the value of the temporary counter 161 exceeds the allowable value), if the temporary counter 161 is operated to be reset, there is some suspicion of vehicle movement. When continuing, the provisional count is reset and the number of steps is not finally counted. Detailed processing of the step count counting unit 160 will be described later.

  The step count display unit 20 displays the number of steps. For example, it can be realized by a liquid crystal display or an organic EL display.

3. Step count processing in this embodiment The step count processing in this embodiment is demonstrated. Specifically, the description will be divided into pre-processing, step count detection processing, feature amount calculation and determination processing, and count processing.

3.1 Pre-processing Pre-processing is performed in the pre-processing unit 110, and is performed by two steps of vector synthesis processing of triaxial acceleration and noise component and DC component removal processing by BPF.

  First, vector synthesis processing of triaxial acceleration will be described. As a sensor signal input from the sensor unit 10, for example, triaxial acceleration information by an acceleration sensor can be considered. Here, the three pieces of acceleration information are not used as they are, but the sum of squares of each axis is calculated and converted into a power vector. Subsequent processing such as step count detection is performed based on this power vector. As the power vector calculation method, the following formula (1), the following formula (2), and the like are conceivable.

  In the above equation (1), the combined waveform may be distorted in a case where the acceleration change due to the vertical movement of walking exceeds ± 1 g. For example, consider a case where most of the components of gravitational acceleration are concentrated on the x-axis and the value is positive. At this time, when the lower peak of fluctuation due to intense vertical motion penetrates to the negative side, the component is folded back positively by the square, and as a result, the waveform is deformed.

  In the above equation (2), the vertical motion component can be more accurately reflected in the power vector by combining the square of the gravitational acceleration component of each axis extracted by the low pass filter (LPF) with a weighting coefficient, and waveform distortion does not occur. Note that LPF (A) in the above equation (2) represents a value obtained by performing low-pass filter processing on the signal A. However, when walking extremely slowly or wearing it in the front pocket of trousers, not only the vertical movement but also the acceleration change (posture change) in the front and rear and right and left contributes to the detection of the number of steps.

  In this embodiment, the vector synthesis method based on the equation (1) is used, but either the above equation (1) or the above equation (2) may be adopted. Another method may be used for calculating the power vector.

  Next, noise component and DC component removal processing by BPF will be described. When the combined power vector of triaxial acceleration is used as it is, it contains high-frequency noise components and DC components as shown in the waveform shown in FIG. 2, and the subsequent step detection processing is appropriately performed. I can't do it. Therefore, the power vector is filtered by BPF to remove high-frequency noise components and DC components that are unrelated to the walking motion. As shown in the BPF (Power) of FIG. 2, the high-frequency noise and the DC component are removed from the signal value after filtering, and a waveform close to a sine wave can be obtained.

3.2 Step Count Detection Process Next, the step count detection process will be described. The step detection process is performed in the step detection unit 120. Specifically, the lower peak detection process by the peak detection unit 121, the evaluation value calculation process by the evaluation value calculation unit 123, and the threshold determination process by the threshold determination unit 125 are executed. Is done.

3.2.1 Lower Peak Detection Processing In a unit cycle of walking motion, vertical acceleration components form a pair in which an upper peak appears first and a lower peak follows. Roughly speaking, the upper peak is when you raise your foot, and the lower peak when you touch the ground. Therefore, the number of steps can be detected by detecting the upper peak or the lower peak.

In this embodiment, the number of steps is detected at the lower peak. Lower peaks are those corresponding to P _i in FIG. 3, for example, specific determination is the following (1), the method (2), prediction and detection to the next step detection obtained from the past walking pitch It is used by adaptively switching according to the period T.
_{(1) P i <P i} -2, when _{P i <P i-1,} P i <= P i + 1, P i < the P _{i + 2,} determined under peak P _i (2) P _i <P _i-3 , P _i <P _i-2 , P _i <P _i-1 , P _i <= P _{i + 1} , P _i <= P _{i + 2} , P _i <= P _{i + 3} It determines P _i under peak judged specifically determined by conditions in (1) of the estimated cycle T is short (approximately 7 or less), (greater than 7) long at conditions in (2).

3.2.2 Process for calculating the value to be evaluated for step count detection The determined lower peak is evaluated as a step count detection candidate to detect the number of steps. The number of steps is detected by threshold determination described later. Here, a method for calculating an evaluated value for threshold determination will be described. In this embodiment, first, the evaluated values R1 and R2 are obtained by the following two methods. The value to be evaluated is a value based on the signal value corresponding to the operation of one step. Since the power vector has the waveform shown in BPF (Power) in FIG. 2 described above, one step of operation is a signal having a waveform corresponding to one cycle of a sine wave. As the value to be evaluated, a value capable of obtaining a waveform characteristic corresponding to one cycle of the sine wave may be employed. For example, an amplitude value, an area (integrated value), or the like may be considered. Here, a value corresponding to the amplitude value is used, but the present invention is not limited to this.

In the first method, a value obtained by inverting the sign of the lower peak is used. Since the DC component is removed after the filter processing by the BPF, the acceleration power vector basically swings up and down around 0. Accordingly, the value to be evaluated is twice the value (-P _i ) obtained by inverting the sign of the lower peak value.
R1 = -2P _i (3)

In the second method, a difference from the maximum value of the past three samples before the lower peak is used. The maximum value of three samples (P _i-2 , P _i-3 , P _i-4 ) before the lower peak is obtained, and a value obtained by subtracting the sample value of the lower peak from the value is set as the evaluated value. The reason why the immediately preceding three samples (P _i-1 , P _i-2 , P _i-3 ) are not used is as follows. As described above, here, the value corresponding to the amplitude value is to be acquired as R2. Therefore, there is an intention to select the maximum value of the past three samples corresponding to the upper peak value. However, it is difficult to assume that the upper peak and the lower peak appear at successive timings from the walking speed (walking pitch) of a person. Therefore, the evaluated value R2 is obtained from the previous three samples that are more likely to correspond to the upper peak value except for the immediately preceding timing _Pi-1 .
R2 = max (P _i-2 , P _i-3 , P _i-4 ) -P _i (4)

  Basically, when the vertical motion by the walking motion is favorably reflected in the acceleration power vector, a sufficient value to be evaluated can be obtained by the method of the above equation (3). However, when the walking pitch is fast, or when the posture changes severely, such as when worn in the front pocket of trousers, the acceleration power vector may not be zero-crossed in synchronization with the walking motion. The number of steps detection becomes unstable with only the value. For example, as shown by the circled portion of the waveform in FIG. 4, the value to be evaluated becomes extremely smaller than the value corresponding to the amplitude value (the value to be evaluated becomes negative when zero crossing is not performed). .

On the other hand, in the value to be evaluated by the method of the above equation (4), when the waveform is close to a sine wave and the amplitude is small (for example, when walking slowly holding the terminal in hand), the value to be evaluated becomes small and the number of steps is detected. It becomes unstable. Therefore, both R1 and R2 are obtained, and the larger one as shown in the following formula (5) is set as the final evaluated value R.
R = max (R1, R2) (5)

3.2.3 Determination of threshold value When the value R to be evaluated obtained by the above equation (5) exceeds a predetermined threshold value, it is determined as the number of steps. Detection of the number of steps by walking motion should normally occur at a certain period. Therefore, in order to distinguish as much as possible from vibrations when riding a vehicle, the determination threshold is not fixed, the next step detection time is predicted, the threshold is lowered near the time when there is a high possibility of detection, and the threshold is set otherwise. Higher is desirable. Therefore, the determination threshold is changed using the prediction detection period T from the previous walking pitch until the next step count detection and the elapsed time t from the previous step count detection as parameters. Specifically, the determination threshold Th is obtained by table lookup using a two-dimensional array as shown in the following formula (6). BaseTh is a variable for raising the entire determination threshold, and is a parameter for adjusting the step count detection sensitivity.
Th = ThLut [T] [t] + BaseTh (6)

  As described above, ThLut only needs to have a characteristic that the threshold value is small at a location where the predicted detection period T and the elapsed time t are close, and the threshold value increases as the difference between T and t increases. The method of setting ThLut is arbitrary. For example, a method of detecting the number of steps using an appropriate number of walking acceleration data with a temporarily set determination threshold and setting based on the distribution of the minimum value of the evaluated values is conceivable. . An example of the determination threshold value table ThLut is shown in FIG. Here, in the example of FIG. 5, BaseTh in the above equation (6) is set to a value of about 0.2, but may be adjusted by the user.

  Note that it is necessary to consider the first step and the second step from the stop state (a state in which the number of steps is not detected exceeding the past one second) as exception processing. When detecting the first step from the stop state, the elapsed time t becomes a very large value as compared with the case where walking is performed continuously, and it is difficult to use the lookup table effectively. Determine using. In addition, since there is no information on the prediction detection cycle T in the subsequent detection of the second step, a dedicated determination threshold table shown in FIG. 6 is used.

Here, a method for obtaining the prediction detection period T will be described. In general, it is better to make prediction based on the closest information in time. Therefore, assuming that the i-th step detection time is t _i , the prediction detection period T _i can be obtained by the following equation (7).

  However, there is a case where the acceleration changes in the right foot step and the left foot step are different when worn in the front pocket of trousers, and the prediction error may increase in the prediction by the above equation (7). FIG. 7 shows an example of the acceleration power vector and the number of steps detected when the step count detecting device 100 is attached to the chest pocket and the user walks normally. When the chest pocket is worn, the change in the acceleration of the body can be observed mainly, so there is almost no difference in the acceleration waveform in the left and right steps, and it can be seen that the detected number of steps is almost equally spaced.

On the other hand, FIG. 8 is an example of wearing a front pocket of pants. When worn in the front pocket of trousers, it can be seen that the step cycle detected is greatly different between right foot → left foot and left foot → right foot due to the influence of foot movement on the wearing side. Therefore, the prediction cycle T _i is a value obtained by subtracting t _i−2 which is the step detection time two steps before the step detection time t _{i−1 one} step before (the following equation (8)).

3.3 Feature Quantity Calculation and Judgment Process Step count information is a powerful information source not only for determining distance traveled, calories burned, and amount of fat burn, but also for determining the user's state. The step count due to must be suppressed as much as possible. Here, a method for suppressing erroneous counting due to vibration during vehicle movement will be described. The feature quantity is extracted by the feature quantity extraction unit 130.

3.3.1 Feature Quantity In the present embodiment, the following three feature quantities are obtained and a composite evaluation is performed to determine vehicle movement.

  As the first feature amount, an acceleration power vector twice-integrated signal (particularly, information on the zero-cross of the twice-integrated signal) is used. A waveform during a general walking motion is shown in FIG. As can be seen from the waveform in FIG. 9, the acceleration signal during the walking motion and the twice-integrated signal have a substantially opposite phase relationship with the same period.

  Next, FIG. 10 shows a waveform due to vibration when the automobile moves. As shown in FIG. 10, the synchronization of the acceleration signal and its twice integrated signal becomes unstable when the vehicle is moving. In particular, when paying attention to the cycle of crossing the zero level, it is greatly different. That is, it is possible to distinguish between normal walking and vehicle movement by comparison processing based on the zero-cross cycle (for example, comparison processing of the number of zero-crosses within a predetermined time).

  However, even when walking, jogging or running with a fast walking pitch, or when worn in the front pocket of trousers, the zero-cross cycle of the acceleration signal and its two-time integration signal do not synchronize like when moving a car. I know that. FIG. 11 shows waveforms of the acceleration power vector and the second-floor seat signal when the number-of-steps detection device 100 is mounted in a pants pocket. However, as can be seen from the vertical scales in FIGS. 10 and 11, in these walking motions, the amplitude of the acceleration power vector is larger than when the vehicle is moving, and the determination based on the magnitude of the evaluated value R in the above equation (5). Thus, it can be distinguished to some extent.

  In summary, first, it is determined that a vehicle whose R is larger than a given threshold value is not a vehicle movement. When R is smaller than a given threshold value, a determination is made based on the zero-cross period of the integral signal twice to distinguish between walking and vehicle movement.

  As the second feature amount, a step period variation (dispersion value) is used. Since the walking motion is a periodic motion, the cycle of the detected number of steps is usually almost constant. On the other hand, vibration during vehicle movement has many random components and large frequency fluctuations. Therefore, the variance of the walking cycle can be obtained, and it can be determined that the vehicle is moving when the variance is large, and is walking when the variance is small. However, even when walking, there may be a large variation immediately after the start of walking, etc., and the evaluation requires caution.

  A noise level is used as the third feature amount. FIG. 12 shows a state of erroneous detection of the number of steps due to vibrations when the automobile is moving. The pulse indicated by the evaluated value R indicates the lower peak detection position of the acceleration power vector, and the height of the pulse is the evaluated value at that time. The pulse indicated by the step count detection indicates the position detected as the step count as a result of the threshold value determination of the evaluated value. As shown in the waveform of FIG. 12, there are some lower peaks that are detected as lower peaks but are finally excluded from the step count detection by the threshold determination (pulses indicated by circles). The ratio of the excluded lower peak is used as a measure for the noise level.

  As long as it is excluded from the detection of the number of steps by the threshold determination, it means that no walking has been performed at that timing. Nevertheless, the detection of the peak value means that the influence of noise is suspected, and here, the evaluated value at the timing of the excluded peak is used as a measure of noise. Specifically, for example, the greater the value to be evaluated at the excluded peak timing, the greater the influence of noise (the possibility of vehicle movement is suspected). However, the magnitude of the signal value varies depending on the exercise state, and it is not preferable to use the value to be evaluated at the excluded peak timing as it is. Therefore, a method such as using a value based on the ratio between the value to be evaluated with the number of steps detected and the value to be evaluated at the excluded peak timing as a feature amount is preferable.

3.3.2 Specific Extraction Method of Feature Quantity In order to extract the three feature quantities described above, it is necessary to define an evaluation period. By using the feature value during the evaluation period, it is determined whether or not the number of steps detected during the evaluation period is vehicle movement. The longer the evaluation period, the more accurate the information. However, as an actual problem, in an evaluation cycle that exceeds the effective number of steps determination period (here 6 seconds), the continuous walking time exceeds the effective number of steps determination period before the feature amount is evaluated, and erroneous detection Cannot be suppressed (a temporary count will be determined before judgment). Therefore, the result must be output within approximately 6 seconds. On the other hand, although there is a method of determining the evaluation period by time, in this embodiment, the evaluation period is determined by the number of steps for the reason that the division when obtaining the variance of the number of steps is omitted. For the above reasons, the evaluation period is 6 steps in this embodiment. Needless to say, the evaluation period is not limited to the period until six steps are detected.

  A description will be given of the twice integrated signal of the acceleration power vector, which is the first feature amount. Here, the number of times of zero crossing of the up-convert signal of the acceleration power vector twice during the detection of the number of steps of 6 steps is counted. If the walking motion is stable, the zero crossing is also 6 times.

  The variance of the step count cycle, which is the second feature amount, will be described. Here, the variance of the step number interval of 6 steps is obtained. However, it is not necessary to accurately calculate the sample variance in the mathematical definition. Since the number of samples is fixed, the pseudo variance without division is obtained. Further, as described in the description of the prediction detection cycle T, the intervals for detecting the number of steps between the right foot and the left foot may not be uniform. Therefore, the variance is also obtained separately by detecting the number of steps on the left and right sides, and finally summed up. Specifically, when the step number interval of 6 steps is T0 to T5, the variance value S is calculated by the following equation (9).

The noise level that is the third feature amount will be described. The total SR of the evaluated values of the six steps detected by the number of steps and the total Sr of all the evaluated values of the lower peak detected while detecting the six steps are obtained. Assuming that the evaluated values for six steps are R _{0 to} R ₅ and the evaluated values excluded in the threshold determination are r ₀ to, the noise level is obtained as Sr from the following equations (10) and (11). However, in practice, values such as the ratio of SR and Sr are used as described above.

3.3.3 Suppression Determination of Step Count Count Each time six steps are detected, the above-described three feature amounts are evaluated, and if it is determined that the vehicle is moving, the effective step count determination period is extended. However, since the effective number of steps determination period cannot be extended indefinitely, when the number of detected steps exceeds 18 steps, for example, when it is determined that the vehicle has moved, the number of steps, walking time, and walking distance up to that point are cleared to zero. .

  Here, in order to extend the effective step number determination period, a time counter for determining whether the effective step number determination period is exceeded is initialized. For example, if the time counter is initialized after 5 seconds from the start of the effective step determination period, the effective step determination period continues for another 6 seconds. That is, the effective step count determination period is extended to a total of 11 seconds. In this method, even if the effective number of steps determination period is extended once, when the next effective number of steps determination period is started, the length returns to the set value (6 seconds) again.

  FIG. 13 shows a flowchart for determining the suppression of the step count. First, the noise level is confirmed in S11. When Sr exceeds Th1 times as much as SR, the clock counter for the effective step count determination period is initialized (S15). The determination of S11 can also be expected to have a suppression effect on vibrations such as vibrators with low vibration amplitude and high frequency.

  If it is determined No in S11, it is confirmed in S12 whether SR is greater than Th2, and the subsequent suppression process is performed only when Th2 or less. Th2 is set so that the number of cases that exceed Th2 is reduced in vehicle movement. Further, in a walking motion in which the value of SR becomes large, the variance value S and the number of zero crossings C of the two-time integration signal also become unstable, and it is highly likely that the number of steps counted by the original walking motion will be affected by the suppression process. Become. Therefore, since it is better not to perform the suppression process in such a case, Th2 may be set from the viewpoint of whether S or C becomes unstable.

  When it is determined No in S12, first, in S13, it is determined whether the number of zero crossings C is six. Here, since the evaluation period is set to the time when 6 steps are detected, C should normally be 6. Therefore, when the value becomes other than 6, the time counter is initialized.

  When it is determined Yes in S13, the dispersion value S is determined in S14. When S exceeds Th3, the time counter is initialized.

  In addition, each threshold value in the flowchart of FIG. 13 may be determined by evaluation based on vibration acceleration data at the time of moving an existing car, train, and walking acceleration data generated by a vibration tester, or may be determined by other methods. Also good.

3.4 Counting Process Next, the step counting process in the step counting unit 160 will be described. The details of the step count processing are shown in the flowchart of FIG. When the processing of FIG. 14 is started, first, in S21, it is determined whether or not a step count detection pulse has been input. If it has not been input, the process ends. If it has been input, the process proceeds to S22. In S22, it is determined whether the continuous walking time is 0, that is, whether it is the first step from the stationary state. If the continuous walking time is 0, the process proceeds to S24.

  If the continuous walking time is not 0 in S22, the effective walking number determination period and the continuous walking time are compared in S23. If the continuous walking time exceeds the effective walking number determination period, the process proceeds to S27. Advances to S25.

  In S25, it is determined whether the effective step number determination period is equal to or greater than the maximum allowable value of the effective step number determination period and the count value of the temporary counter 161 is equal to or greater than the maximum allowable value. If Yes, the process proceeds to S27, and if No, the process proceeds to S26.

  S26, S27, and S24 are count processes for changing one or both of the temporary counter 161 and the final output counter 163. Specifically, in S26, the count value of the temporary counter 161 is incremented. This is not the case at the start of walking, and the case where the effective step count determination period and the counter value of the temporary counter 161 are within the maximum allowable value. That is, it corresponds to the continuation of provisional count processing.

  In S27, the count value of the temporary counter 161 is added to the count value of the final output counter 163. This corresponds to processing for determining and confirming that the number of steps tentatively counted by using the temporary counter 161 is due to walking. Then, the final output counter 163 is incremented in correspondence with the number of steps corresponding to the number of steps detection pulse input in S21, and the temporary counter 161 is reset.

  In S24, the effective number of steps determination period is set to an initial value, and 1 is set in the temporary counter 161. This corresponds to a case where walking is started, and is a process of starting a temporary count using the temporary counter 161.

  If any of the processes in S24, S26, and S27 ends, it is determined in S28 whether the effective step count determination period is smaller than the maximum allowable value and whether a time counter reset command for the effective step count determination period is input. . Here, as described above, the time counter reset command is input to the step counting unit 160 as a result of the determination by the determination unit 140. In the case of Yes, the time counter is reset in S29. If it is determined No in S28, or if the process of S29 is completed, this process ends.

  In the above embodiment, as shown in FIG. 1, the step detection device 100 includes the step detection unit 120 that detects the number of steps based on the sensor signal from the acceleration sensor, and the walking that measures the continuous walking time from the start of walking. A time measurement unit 150, a step count unit 160 that performs step count processing based on the result of step detection by the step detection unit 120, and a feature amount for determining whether or not a vehicle is moving A feature amount extraction unit 130 that extracts based on the signal and a determination unit 140 that determines whether the user is moving the vehicle based on the extracted feature amount. Then, when the continuous walking time is shorter than a given effective number of steps determination period, the step count counting unit 160 resets the count of the number of steps detected during the continuous walking time. At the same time, the step counting unit 160 performs a process of extending the effective step determination period when it is determined that the user is moving based on the determination result of the determination unit 140.

  Here, the continuous walking time is a time during which walking is continuously performed. The continuous walking represents a state in which the number of steps is detected next during a given time from the previous number of steps detection. Further, the effective step count determination period is a period for determining whether the value counted by the temporary counter 161 is added to the final output counter 163 or discarded as described in the system configuration section. The count is reserved by extending.

  The relationship between the sensor unit 10 having an acceleration sensor, the step count detection unit 120, and the feature amount extraction unit 130 is, for example, as illustrated in FIG. The sensor unit 10 is connected to the preprocessing unit 110, and DC components and the like are removed by performing preprocessing. In FIG. 1, the sensor unit 10 and the preprocessing unit 110 are described as being different from each other, but the preprocessing unit 110 may be included in the sensor unit 10. In this case, since the sensor unit 10 performs processing on the sensor signal, the output from the sensor unit 10 is not the sensor signal itself but a signal after processing. The preprocessing unit 110 is connected to the step count detection unit 120 and the feature amount extraction unit 130. The feature amount extraction unit 130 extracts the feature amount from the signal (acceleration power vector) from the preprocessing unit 110, the evaluation value calculated by the evaluation value calculation unit 123, and the threshold value determination unit 125. A step detection pulse obtained as a result of the threshold determination is also used. In FIG. 1, the connection between the step count detection unit 120 and the feature amount extraction unit 130 is described with a single arrow for simplicity, but a plurality of types of signals are output as described above.

  As a result, when it is determined that the user is moving the vehicle, the effective number of steps determination period is extended. Therefore, the number of steps detected when the vehicle is determined to be moved is retained or discarded. Thus, erroneous counting due to vehicle movement can be suppressed. The reason why the erroneous count can be suppressed by extending the effective step count determination period is as described above.

  In addition, the number-of-steps detection unit 120 obtains an evaluation value for walking based on a sensor signal at a timing specified by a timing at which a peak value is detected. Then, the determination unit 140 determines that the user is not moving the vehicle when the evaluated value at the timing determined as walking among the evaluated values in the evaluation period is larger than a given threshold.

  Here, the peak value is detected by the peak detection unit 121 of the step count detection unit 120, and the actual processing is to obtain the above-described equations (3) to (5). In the present embodiment, the value to be evaluated is calculated at the timing when the peak is detected, but the value to be evaluated may be calculated at other timings. After the evaluation value is calculated, threshold determination is performed by the threshold determination unit 125. As a result, there may be an evaluation value that is excluded as not being a walk. Therefore, the evaluated value includes an evaluated value determined as walking and an evaluated value not determined as walking. The evaluation period is a period for determining whether or not the detected number of steps is due to vehicle movement based on the feature amount. As described above, since it is necessary to make a determination before the continuous walking time exceeds the effective number of steps determination period, the evaluation period is set to be shorter than the effective number of steps determination period. In this embodiment, for example, the number of steps is detected six times. It is a period to do. Therefore, there are usually six (6) timing-evaluated values determined as walking during the evaluation period. When comparing this with a threshold value, for example, the sum of six evaluated values may be used, or an average may be used.

  Thereby, based on the evaluated value determined to be walking, when the value is larger than a given threshold, it can be considered to be walking. Therefore, it is not necessary to make a determination using other feature amounts, and it is possible to easily determine whether or not the vehicle is moving. This is because, as shown in FIGS. 10 and 11, it is difficult to consider a movement with a signal value larger than a certain level as a vehicle movement. Of course, there may be a walking motion with a small signal value, but in this case, the determination may be made using other feature amounts.

  The feature amount extraction unit 130 may extract the number of zero crosses of the twice integrated signal value of the acceleration power vector as the feature amount. The determining unit 140 compares the number of steps detected by the number of steps detecting unit 120 with the number of zero crosses when it is determined that the value to be evaluated determined to be walking in the evaluation period is smaller than a given threshold value. Then, it is determined whether or not the vehicle is moving.

  Here, the number of zero crossings refers to the number of times that the signal value becomes zero in a given period.

  This makes it possible to distinguish between walking and vehicle movement even when it is not possible to determine walking by the magnitude of the value to be evaluated. This is because, as shown in FIG. 9, in the case of walking, the acceleration power vector and its twice integrated signal value have the same cycle in opposite phases, but such characteristics do not appear in vehicle movement. . Specifically, if the evaluation period is a time for detecting 6 steps, the number of steps detected by the number-of-steps detection unit 120 is 6. Therefore, it is determined whether the number of zero crossings is 6.

  The feature amount extraction unit 130 may extract the number of zero crosses of the twice integrated signal value of the acceleration power vector as the feature amount. The determining unit 140 compares the number of steps detected by the number of steps detecting unit 120 with the number of zero crossings to determine whether the vehicle is moving.

  This makes it possible to distinguish between the normal walking shown in FIG. 9 and the vehicle movement shown in FIG. However, it should be noted that it is necessary to distinguish the case where the number-of-steps detection device 100 is mounted in a pants pocket using other feature amounts (FIG. 11) from vehicle movement.

  In addition, the number-of-steps detection unit 120 may obtain an evaluation value of walking based on a sensor signal at a timing specified by a timing at which a peak value is detected. Then, the determination unit 140 may determine whether the user is moving based on the relationship between the evaluated value determined to be walking and the evaluated value determined not to be walking. Specifically, a ratio between a value obtained based on the sum of evaluated values determined to be walking and a value obtained based on the sum of evaluated values determined not to walk may be calculated. Then, the calculated ratio is compared with a given threshold value.

  This makes it possible to determine whether the vehicle is moving by determining the noise level. As shown in FIG. 12, the evaluated value may include an evaluated value determined as walking and an evaluated value determined as not walking. Normally, the peak value is detected only during walking, and should not be detected anywhere other than walking. Nevertheless, it is necessary to doubt other factors such as noise that the peak value has been detected. Here, when the value of the value to be evaluated determined not to be walking is large (as described above, it is desirable to use the ratio instead of the value as it is), it is considered that the influence of noise is large. When the influence of noise is large, a count reservation process is performed.

  Further, the step count counting unit 160 includes a temporary counter 161 and a final output counter 163. Then, until the continuous walking time exceeds the effective number of steps determination period, the number of steps is counted by incrementing the temporary counter 161. When the continuous walking time exceeds the effective number of steps determination period, the value of the temporary counter 161 is set. Add to the final output counter 163.

  Here, the number of steps presented to the user is the count value of the final output counter 163, and the value of the temporary counter 161 is not presented to the user. However, it does not prevent the user from referring to the temporary counter by a specific operation.

  This makes it possible to perform a counting process using two counters, the temporary counter 161 and the final output counter 163. The reset process when the continuous walking time is shorter than the effective number of steps determination period described above corresponds to the process of resetting the value of the temporary counter 161 without adding it to the final output counter 163.

  The step count detection unit 120 may set a threshold based on the prediction detection cycle and the elapsed time since the previous step count detection, and may perform the step count detection process when the evaluated value is larger than the threshold. Specifically, a lookup table in which an output value is determined based on the prediction detection cycle and the elapsed time from the previous step count detection may be provided, and the output value of the lookup table may be set as a threshold value.

  As a result, it is possible to adaptively set the threshold according to the prediction detection cycle and the elapsed time since the previous step detection. Specifically, a method using the lookup table shown in FIG. When a person walks, the walking cycle is usually almost constant or changes slowly, so once the predicted detection cycle is set, the next step counts for the predicted detection cycle from the previous step count detection. Is likely to be detected when Therefore, the threshold value is set low at such timing (diagonal component of the look-up table), and the threshold value is set high at timing when the cycle changes suddenly.

  In the above-described embodiment, the step detection unit 120 that detects the number of steps based on the sensor signal from the acceleration sensor, and the step count unit 160 that performs the step count processing based on the result of the step detection by the step detection unit 120. Is related to the step detection device 100 including The step count detection unit 120 sets a threshold value based on the prediction detection cycle and the elapsed time since the previous step count detection, and performs the step count detection process when the evaluated value is larger than the threshold value. Also in this step count detection apparatus 100, a lookup table in which an output value is determined based on a prediction detection cycle and an elapsed time since the previous step count detection may be used.

  This makes it possible to realize a step count detection apparatus that adaptively sets the threshold value.

Further, the step detection unit 120, when detecting the step detection time _t 1 ~t _N first to N, the predicted detection period _{T N} of the N used in the (N + 1) of the step detection process _T N = t You may obtain | require as _N-1- tN _-2 .

  This makes it possible to set an appropriate prediction detection cycle even when the step detection device 100 is worn in a pants pocket. When worn in the pocket of the pants, as shown in FIG. 8, a long cycle and a short cycle appear alternately at the timing of putting out the right foot and the timing of putting out the left foot. Therefore, instead of using the immediately preceding two step count detection times as shown in Expression (7), it is preferable to use the previous timing as shown in Expression (8).

  Further, the present embodiment described above relates to an electronic device including the step count detection apparatus 100 described above.

  Thereby, it is possible to realize an electronic device in which the step detection device 100 is built. Specifically, as shown in FIG. 15, it is assumed to be built in a portable electronic device such as a cellular phone. The portable electronic device includes a sensor unit 10, a step count detection device 100, a communication unit 210, a processing unit 220, a display unit 230, an I / F unit 240, and an information storage medium 250. However, the portable electronic device is not limited to the configuration of FIG. 15, and various modifications such as omitting some of these components or adding other components are possible. For example, since the portable electronic device has the display unit 230, the step number display unit 20 shown in the pedometer of FIG. In addition, although the step count detection apparatus 100 and the processing unit 220 are different from each other, the processing performed by the step count detection unit 120, the step count counting unit 160, and the like may be performed by the processing unit 220 of the portable electronic device. For example, when the step count detection apparatus 100 is realized by software, it is conceivable that the step count detection apparatus 100 is implemented in an electronic device as a program that executes processing of each unit of the step count detection apparatus 100 in FIG. Specifically, the program is stored in the information storage medium 250 shown in FIG. 15 and connected to the electronic device via the I / F unit 240. Then, the processing unit 220 reads the program from the information storage medium 250 and performs the step count detection process by executing the process described in the program. In that case, since it is not necessary to perform the step detection process in the block of the step detection device 100 shown in FIG.

  Further, in the present embodiment described above, the number of steps detecting unit 120 that detects the number of steps based on the sensor signal from the acceleration sensor, the walking time measuring unit 150 that measures the continuous walking time from the start of walking, and the number of steps detecting unit 120 A step count counting unit 160 that performs step count processing based on the result of step detection, and a feature amount extraction unit 130 that extracts a feature amount for determining whether or not the vehicle is moving based on the sensor signal. And a determination unit 140 that determines whether or not the user is moving based on the extracted feature amount, and relates to a program that causes a computer to function. Then, when the continuous walking time is shorter than a given effective number of steps determination period, the step count counting unit 160 resets the count of the number of steps detected during the continuous walking time. At the same time, the step counting unit 160 performs a process of extending the effective step determination period when it is determined that the user is moving based on the determination result of the determination unit 140.

  Further, in the present embodiment, the number of steps detecting unit 120 that detects the number of steps based on the sensor signal from the acceleration sensor, and the number of steps counting unit that performs the step counting process based on the result of the number of steps detection by the number of steps detecting unit 120. As 160, the computer functions and is related to the cocoon program. The step detection unit 120 sets a threshold based on the step detection detection cycle and the elapsed time since the previous step detection, and based on the sensor signal at the timing specified by the timing at which the peak value is detected. The number of steps is detected when the obtained evaluated value is larger than the threshold value.

  As a result, it is possible to realize a program that performs processing for counting the number of steps in software, not in hardware. The program is recorded on an information storage medium. Here, as the information recording medium, various recordings that can be read by the step count detection apparatus 100 and electronic devices such as optical disks such as DVD and CD, magneto-optical disks, hard disks (HDD), memories such as nonvolatile memory and RAM, and the like A medium can be assumed. For example, as described above with reference to FIG. 15, a case where the program is stored in the information storage medium 250 of the portable electronic device and executed in the processing unit 220 can be considered. More specifically, it is assumed to be realized as a mobile phone application. Some mobile phones in recent years are equipped with a triaxial acceleration sensor to detect the posture of the mobile phone. Therefore, by downloading the pedometer application from the network via the communication unit 210 and acquiring the program and installing the acquired program, the mobile phone can be made to perform the step detection process. In this case, the processing unit 220 specifically refers to a CPU of a mobile phone. The information storage medium 250 may be a memory on a chip mounted inside the mobile phone, or may be a card-type memory that is used by being inserted into a slot provided in the mobile phone.

  Although the present embodiment has been described in detail as described above, it will be easily understood by those skilled in the art that many modifications can be made without departing from the novel matters and effects of the present invention. Accordingly, all such modifications are intended to be included in the scope of the present invention. For example, a term described at least once together with a different term having a broader meaning or the same meaning in the specification or the drawings can be replaced with the different term in any part of the specification or the drawings. Further, the configuration and operation of the step detection device, the electronic device, and the like are not limited to those described in this embodiment, and various modifications can be made.

10 sensor units, 20 step number display units, 110 preprocessing units, 120 step number detection units,
121 peak detection unit, 123 evaluation value calculation unit, 125 threshold determination unit,
130 feature amount extraction unit, 140 determination unit, 150 walking time measurement unit,
160 step count section, 161 temporary counter,
163 Final output counter, 210 communication unit, 220 processing unit, 230 display unit

Claims (14)

A step detection unit that detects the number of steps based on a sensor signal from the acceleration sensor;
A walking time measurement unit that measures the continuous walking time from the start of walking; and
A step counting unit that performs a step counting process based on a result of the step detection in the step detecting unit;
A feature amount extraction unit for extracting a feature amount for determining whether or not the vehicle is moving based on the sensor signal;
A determination unit that determines whether the user is moving a vehicle based on the feature amount extracted by the feature amount extraction unit;
Including
The step count unit
If the continuous walking time is shorter than a given effective number of steps determination period, reset the count of steps detected in the continuous walking time,
The step count unit
A step detection device that extends the effective step determination period when it is determined that a user is moving based on a determination result in the determination unit. In claim 1,
The step detection unit
Obtaining an evaluation value of the walking based on the sensor signal at a timing specified by a timing at which a peak value is detected;
The determination unit
An apparatus for detecting the number of steps, wherein, when an evaluated value determined to be walking among the evaluated values in an evaluation period is larger than a given threshold value, it is determined that the user has not moved the vehicle. In claim 2,
The feature amount extraction unit includes:
The number of zero crossings of the twice integrated signal value of the acceleration power vector is extracted as the feature amount,
The determination unit
When it is determined that the evaluated value determined as walking among the evaluated values in the evaluation period is smaller than a given threshold, the number of steps detected by the number of steps detection unit and the number of zero crosses And determining whether or not the user is moving the vehicle based on the comparison result. In claim 1,
The feature amount extraction unit includes:
The number of zero crossings of the twice integrated signal value of the acceleration power vector is extracted as the feature amount,
The determination unit
A step count detection apparatus that compares the number of steps detected by the step count detection unit with the number of zero crossings and determines whether the user is moving on the basis of the comparison result. In claim 1,
The step detection unit
Obtaining an evaluation value of the walking based on the sensor signal at a timing specified by a timing at which a peak value is detected;
The determination unit
A determination is made as to whether the user is moving on the basis of the relationship between the evaluated value determined as walking among the evaluated values and the evaluated value determined as not walking. Step detection device. In claim 5,
The determination unit
Of the evaluated values in the evaluation period, a ratio between a value obtained based on the sum of the evaluated values determined to be walking and a value obtained based on the sum of the evaluated values determined not to be walking And determining whether the user is moving the vehicle by comparing the obtained ratio with a given threshold value. In any one of Claims 1 thru | or 6.
The step count unit has a temporary counter and a final output counter,
The step count unit
When the continuous walking time is shorter than the effective number of steps determination period, the temporary counter performs the step counting process, and when the continuous walking time exceeds the effective number of steps determination period, the value of the temporary counter is set. A step detection device that performs a process of adding to a value of a final output counter. In any one of Claims 1 thru | or 7,
The step detection unit
A threshold value is set based on the predicted number of steps detection period and the elapsed time since the previous step number detection, and the evaluated value obtained based on the sensor signal at the timing specified by the timing at which the peak value is detected The step count detection device performs the step count detection when the threshold is larger than the threshold value. A step detection unit that detects the number of steps based on a sensor signal from the acceleration sensor;
A step counting unit that performs a step counting process based on a result of the step detection in the step detecting unit;
Including
The step detection unit
A threshold value is set based on the predicted number of steps detection period and the elapsed time since the previous step number detection, and the evaluated value obtained based on the sensor signal at the timing specified by the timing at which the peak value is detected The step count detection device performs the step count detection when the threshold is larger than the threshold value. In claim 8 or 9,
The step detection unit
A step number in which an output value is determined based on the prediction detection cycle and the elapsed time from the previous step number detection, and the output value of the look-up table is set as the threshold value. Detection device. In any one of Claims 8 thru | or 10.
The step detection unit
When detecting step detection time _t 1 ~t _N first to N, the predicted detection period _{T N} of the N used in the (N + 1) of the step detection process _{T N = t N-1 -t} N-2 A step count detection apparatus characterized by obtaining as follows.   An electronic apparatus comprising the step detection device according to claim 1. A step detection unit that detects the number of steps based on a sensor signal from the acceleration sensor;
A walking time measurement unit that measures the continuous walking time from the start of walking; and
A step counting unit that performs a step counting process based on a result of the step detection in the step detecting unit;
A feature amount extraction unit for extracting a feature amount for determining whether or not the vehicle is moving based on the sensor signal;
Based on the feature amount extracted by the feature amount extraction unit, as a determination unit that determines whether the user is moving the vehicle,
Make the computer work,
The step count unit
If the continuous walking time is shorter than a given effective number of steps determination period, reset the count of steps detected in the continuous walking time,
The step count unit
A program for extending the effective step count determination period when it is determined that the user is moving based on a determination result in the determination unit. A step detection unit that detects the number of steps based on a sensor signal from the acceleration sensor;
As a step count unit that performs a step count process based on the result of the step detection by the step detector,
Make the computer work,
The step detection unit
A threshold value is set based on the predicted number of steps detection period and the elapsed time since the previous step number detection, and the evaluated value obtained based on the sensor signal at the timing specified by the timing at which the peak value is detected When the number of steps is larger than the threshold, the number of steps is detected.

Images