JP4894500B2 - Walking waveform processing method and walking waveform processing apparatus - Google Patents

Description

  The present invention relates to a walking waveform processing method and a walking waveform processing apparatus, and is suitable for application to, for example, detecting the number of steps.

  Conventionally, a method of identifying a one-step waveform corresponding to one step of a walk, using an amplitude peak that appears specifically as an index among displacements of an electric field (walking waveform) formed in the human body as the human body walks Has already been proposed by the present applicant (see Patent Document 1).

  Such a specific amplitude peak appears immediately after the toe completely leaves. Immediately after the toe of the right foot (left foot) completely leaves the ground, the left foot (right foot) will be in a completely landing state regardless of the walking mode. Since no charging interference occurs, the largest amplitude peak in the walking waveform appears almost uniformly in the band of 8 [Hz] ± 2 [Hz] ((2) and (4) in FIG. 6).

By detecting a peak of amplitude appearing in the band of 8 Hz ± 2 Hz from the walking waveform, and using this as a reference for the one-step waveform corresponding to one step, it is possible to accurately extract the one-step waveform.
JP 2004-147793 A

  However, as a result of subsequent experiments, the timing of the right toe (left foot) toe and the timing to touch the left foot (right foot) heel is close, so the amplitude peak corresponding to the toe's detachment is It has been found that it may not appear due to interference with the peak of amplitude corresponding to grounding.

  Therefore, if the amplitude peak appearing in the band of 8 Hz ± 2 Hz is used as a reference for a one-step waveform corresponding to one step, there is a problem that one step in walking cannot be specified accurately.

  The present invention has been made in consideration of the above points, and intends to propose a walking waveform processing method and a walking waveform processing apparatus capable of accurately detecting walking.

  In order to solve such a problem, the present invention is a step waveform processing method, wherein electric field displacement in a frequency band of 20 [Hz] or less formed around a human body, from positive and negative polarities based on a DC component, The amplitude peak of one of the set polarities is searched, and the maximum amplitude peak having the same polarity as the polarity of the amplitude peak in the predetermined time width centered on the searched amplitude peak is set in the swing phase. Detection was made as the resulting amplitude peak.

  The present invention is also a walking waveform processing apparatus, a detection unit that detects an electric field, an extraction unit that extracts a signal component in a frequency band of 20 [Hz] or less from the electric field detected by the detection unit, and an extraction unit A peak detection unit that detects an amplitude peak that occurs in the swing phase from the signal component extracted by the signal component, and the peak detection unit, from the signal component, out of positive and negative polarities based on the DC component in the signal component, An amplitude peak having one of the set polarities is searched, and a maximum amplitude peak having the same polarity as the polarity of the amplitude peak is generated in the swing phase during a predetermined time width centered on the searched amplitude peak. Detection was made as an amplitude peak.

  As described above, according to the present invention, one appears for each step, and one corresponding to appear when one foot is most separated from the road surface and the other leg is in a stable state, That is, the peak at which the influence of the interference between the component caused by the free leg and the component caused by the standing foot can be used as a reference for one step. It can count more accurately.

  In addition, the searched amplitude peak is not immediately set as the amplitude peak of the swing phase, but the maximum amplitude peak having the same polarity as the polarity of the amplitude peak within the predetermined time width centered on the amplitude peak is Since the amplitude peak is used, an amplitude peak caused by a potential change (noise) that does not correspond to walking can be eliminated. As a result, for example, the number of steps can be counted more accurately.

(1) Relationship between charging of human body by walking and electric field around human body It is well known that human body is charged by walking. This human body voltage can be found in the literature ([Creator] Hirokazu Kimura, [Title] Chargeability of fiber / polymer material and its evaluation method, [Media type] online, [Issue date] June 29, 1998. [Source] Industrial Technology Textile Group, Evaluation Technology Department, [Source Address] http://www.tri.pref.osaka.jp/group/sense/sangyoseni/static.ele.pdf) It is strongly dependent on the surface material) and the type of footwear, and the combination of the rug and footwear has the lowest absolute value of the human body voltage at about 480 [V].

  The electric field around the human body is simulated when the floor is at a reference potential of 0 [V] and the human body is charged to 480 [V], and the distribution pattern of the electric field strength obtained from the simulation is shown in FIG. As a condition for the simulation, a human body model in an upright posture having a dielectric constant of a living tissue shown in FIG. 2 exists in a space where there is only a floor without walls and ceilings. Incidentally, the simulation software calculates the electric field in the three-dimensional space using the Poisson equation.

  As is clear from FIG. 1, it can be seen that the electric field is distributed around the human body. Since the electric field vector is in the normal direction of the human body surface, if the human body surface and an electrode spaced in the normal direction from the human body surface are arranged and the potential difference between the pair of electrodes is detected, It becomes possible to obtain a change in the charged voltage of the human body due to walking. This potential difference is generally proportional to the charged voltage of the human body, and is approximately 0.02 times the charged voltage.

For example, when the placement target of the pair of electrodes is a wrist, the electric field strength at the wrist is about 1000 [V / m], and if an electrode with an interelectrode distance of 1 [cm] is placed on the wrist, The potential difference between the electrodes is approximately 10 [V]. The theoretical noise floor in this case is -154 [dB] assuming that the electrode area is 1 [cm] x 1 [cm], the noise coefficient is 10, and the frequency band of the walking waveform is 10 [Hz]. When converted into an electric field, it becomes approximately 6 × 10 ⁻⁹ [V / m]. Therefore, under such conditions, if the potential change of the human body due to walking is 6 × 10 ⁻⁹ [V / m] or more, the change can be detected.

(2) Relationship between walking mode and walking waveform As shown in FIG. 3, the walking mode on the right foot is immediately after the heel has taken off (referred to as leaving from the road surface; the same applies hereinafter) and immediately before the toes off. The ground removal process (FIG. 3 (A)), the kicking process (FIG. 3 (B)) from immediately after the toe has just landed to just before the heel has landed, and the entire sole from immediately after the heel has landed. Three types of processes are sequentially repeated, including a landing process (FIG. 3C) until landing on the road surface (hereinafter referred to as complete landing).

  As for the human body walking mode on the left foot, the three types of processes are repeated in the same way as the right foot, but the start time of each process for the left foot is different from that for the right foot, and the landing process occurs during the right foot takeoff process. A process is started (indicated by an arrow in FIG. 3A), and a takeoff process in the left foot is started in the middle of the landing process in the right foot (indicated by an arrow in FIG. 3C).

  In this way, the walking of the human body is caused by the left foot landing process, which is opposite to the right foot landing process, and the left foot landing process, which is opposite to the right foot landing process, approximately half a cycle apart, It is repeated alternately.

  On the other hand, the potential change (walking waveform) formed around the human body by walking is equipotential because the human body has high electrical conductivity and dielectric constant, and mainly the charge amount on the sole of the foot (that is, the charge amount of the human body). This corresponds to a change in capacitance between the bottom surface of the foot and the road surface.

  Here, when the potential of the human body is “V”, the charge amount of the human body is “Q”, and the capacitance of the human body is “C”, the potential of the human body can be expressed as “V = Q / C”. . The charge amount Q of the human body may be either positive or negative depending on the relationship between the charge series caused by the material of the road surface and the sole (sole).

  For example, the charge amount Q of the human body is positive under the condition that the material of the road surface is soft PVC and the sole is skin. FIG. 4 shows a potential change (walking waveform) caused by one leg under this condition.

  As can be seen from FIG. 4, the walking waveform resulting from one step of one foot has several characteristic amplitude peaks. First, at the time when the kite leaves (corresponding to the second from the left in FIG. 3A), the charge amount Q increases (peeling charging) with rapid peeling of the kite from the road surface, and the capacitance C Decreases, a negative amplitude peak (hereinafter referred to as an H wave) appears.

  After that, even when the bottom of the foot is gradually separated from the road surface and the toe is released (corresponding to the right end in FIG. 3 (A)), the toe is rapidly separated from the road surface as in the case of the generation of the H wave. Along with this, the charge amount Q increases (peeling charge) and the capacitance C decreases, so that a negative amplitude peak (hereinafter referred to as an I wave) appears.

  Eventually, when the foot away from the road surface starts to accelerate and the acceleration of the foot becomes the fastest (corresponding to the second from the left in FIG. 3B (called the acceleration period)), the position of the foot relative to the road surface is the most. Since the capacitance between the road surface and the foot is minimized, a positive amplitude peak (hereinafter referred to as a J wave) appears.

  Thereafter, a gentle positive amplitude peak (hereinafter referred to as a K wave) appears when the foot is swung in the traveling direction, and the swung foot begins to decelerate. Amplitude peak (hereinafter referred to as L wave) appears.

  Eventually, when the decelerated foot heel touches the road surface (corresponding to the left end in FIG. 3C), the charge amount Q decreases as the heel leaves the road surface (electric charge is discharged from the sole), When the capacitance C increases, a positive amplitude peak (hereinafter referred to as an M wave) appears, and when the sole completely contacts the road surface (the right end in FIG. 3C). Correspond to the road surface, there is no take-off area of the bottom of the foot with respect to the road surface, so a positive amplitude peak (hereinafter referred to as N wave) appears.

  The period from the I wave to the M wave is called the swing phase because the foot is completely lifted off the road surface, and the period from the M wave to the next I wave is the road surface. It is called the stance phase because it has arrived.

  On the other hand, FIG. 5 (A) shows a potential change (walking waveform) caused by both feet under the same conditions as FIG. The walking waveform based on both feet differs greatly from the walking waveform based on one foot (Fig. 4) due to a half-cycle shift in the walking process of the right and left feet, but the ground movement and ground contact, which are the basic motions of walking. In this case, characteristic peaks (a portion indicated by a solid line (corresponding to a J wave) and a portion indicated by a broken line (corresponding to an I wave and an M wave) in FIG. 5A) appear. By the way, in the walking waveform of one foot (Fig. 4), the M wave (N wave) is higher than the J wave, but in the walking waveform of both feet, the M wave (N wave) interferes with the I wave, etc. Is lower than the J wave.

  Among these characteristic peaks, in Japanese Patent Application Laid-Open No. 2004-147793 already disclosed by the present applicant, the left foot (right foot) is completely landed when the right foot (left foot) toe completely leaves the ground. The negative amplitude peak (I wave) that appears corresponding to 1 appears at about 8 [Hz] ± 2 [Hz] even if both feet are moving alternately with a shift of about a half cycle. It was an indicator.

  However, the amplitude peak (I wave) corresponding to the toe separation on one foot is in a state where the amplitude is low, as is apparent from the potential change (walking waveform) caused by both feet shown in FIG. (A portion indicated by a broken line in FIG. 5B (corresponding to I wave and M wave)) may appear. This is mainly because the timing at which the toe of the right foot (left foot) leaves and the timing at which the heel of the left foot (right foot) contacts are close to each other, so that both amplitude peaks interfere with each other. Incidentally, the walking waveform shown in FIG. 5B is also under the same conditions as in FIG.

  Therefore, in the free leg period (period from the I wave to the M wave) in which one leg is completely lifted from the road surface and is standing on the other leg, it is 2 from the left in the acceleration period (FIG. 3B). J wave appearing in the second) is adopted as a detection index for one step.

  The reason for adopting this J-wave is that the acceleration period is when one foot is farthest from the road surface. At this time, the stance of the other foot is in a particularly stable state, so the waveform component at this time is This is because the component caused by the free leg is mainly, that is, the interference between the component caused by the free leg and the component caused by the leg standing is the smallest.

  Therefore, in the present invention, of the displacement of the electric field formed in the human body due to the biped movement of the human body, the amplitude peak that occurs when one foot is farthest from the road surface, that is, during the swing phase, is used as an index. It will be.

  4 and 5, the case where the J wave appears as a positive amplitude peak and the I wave appears as a negative amplitude peak is illustrated. However, as described above, the road surface and the sole (sole) In some cases, the positive and negative polarities of the amplitudes may be reversed depending on the relationship of the charging series in FIG. 6, so that the J wave may appear as a negative amplitude peak and the I wave may appear as a positive amplitude peak.

(3) Embodiment Hereinafter, an embodiment of the present invention will be described in detail.

(3-1) Overall Configuration of Pedometer FIG. 6 shows the overall configuration of the pedometer according to the present embodiment. The pedometer 1 is configured by connecting a wristband 3 to a rectangular main body 2. As shown in FIG. 7, the surface of the body 2 facing the human body surface (hereinafter referred to as the skin-facing surface) KTa keeps the distance between the skin-facing surface and the human body constant. For this reason, an error on the pedometer side based on a change in distance caused by vibration or the like can be avoided in advance.

  Further, the main body 2 is provided with a first electrode (hereinafter referred to as a human body side electrode) E1 on the back surface of the skin facing surface KTa, and the second electrode on the back surface of the surface facing the skin facing surface KTa. E2 (hereinafter referred to as the outer electrode) is provided.

  In this pedometer 1, when the main body 2 is attached to the wrist by the wristband 3, the main body 2 takes in a potential difference between the human body side electrode E1 and the outer electrode E2 as a potential change of the human body, and based on the potential change. The number of steps is counted.

(3-2) Electrical Equivalent Circuit of Human Body and Pedometer Next, FIG. 8 shows an electrical equivalent circuit of the human body and the main body 2 when the main body 2 is attached to the wrist of the human body.

  In this case, the human body is coupled to the road surface (ground) (C1) and coupled to the outer electrode E2 (C2). Further, since the outer electrode E2 is coupled (C3) to a ground such as a floor or a wall, it becomes a virtual ground.

  As can be seen from this equivalent circuit, the human body capacitance (C4), the internal resistance (R2) in the conductive pad PD, and the internal resistance (R3) in the main body 2 are approximately fixed. The potential difference generated with the outer electrode E2 can be regarded as a change in the potential of the human body.

(3-3) Circuit Configuration of Main Body Next, a circuit configuration of the main body 2 will be described. As shown in FIG. 9, the main body 2 amplifies the potential difference between the outer electrode E2 and the human body side electrode E1 by an amplifier 11, and sends the amplified result to a BPF (Band Pass Filter) 12 as a human body potential signal.

  The BPF 12 extracts a range of 20 [Hz] from the human body potential signal with reference to the direct current component as a component corresponding to an electric field formed around the human body in accordance with the walking of the human body, and extracts the extracted signal (hereinafter referred to as the signal). , Which is called a walking waveform signal) is sent to an A / D (Analog / Digital) converter 13. The A / D converter 13 performs A / D conversion processing on the walking waveform signal, and sends human body potential data D1 obtained as a result of the processing to the processor 14.

  The processor 14 has a computer configuration including a CPU (Central Processing Unit), a ROM (Read Only Memory) in which various programs are stored, a RAM (Random Access Memory) as a work memory, a counter, and a timer. Based on the stored program, the step counting process is executed. The step count processing of the processor 14 is specifically performed according to the flowchart shown in FIG.

  That is, when a signal indicating a counting start request is given from the operation unit (not shown), the processor 14 starts the step count processing procedure RT after setting an internal counter to an initial value. One of the positive and negative polarities based on the component is initialized as an amplitude peak search target.

  Next, in step SP2, the processor 14 searches for the amplitude peak of one of the initially set polarities from the walking waveform (eg, FIG. 5A) of the human body potential data supplied from the A / D converter 13, The searched amplitude peak is set as a candidate for an amplitude peak (J wave) generated during the swing phase.

  In step SP3, the processor 14 sets a window having a predetermined time width around the searched amplitude peak, and in the subsequent step SP4, the processor 14 sets the same polarity as one of the initial polarities. Is detected as a J wave.

  The processor 14 proceeds to step SP5, stores the appearance time of the current detection target J wave (hereinafter referred to as the current J wave) in the memory 15 as data, and then in the next step SP6, the current J wave. Between the wave and the J wave detected as the current detection target immediately before the current J wave (hereinafter referred to as the immediately preceding J wave), the polarity of one of the initially set polarities is reversed. It is determined whether the number of appearances of the amplitude peak of the other polarity is 1 or less.

  Here, when two or more amplitude peaks having opposite polarities to those of the J wave appear between the current J wave and the immediately preceding J wave, this means that the amplitude peak detected as the J wave in step SP4. Is likely to be an I wave, that is, it is highly likely that an I wave is being measured. This can be easily understood by looking upside down in the graph in FIG.

  In this case, the processor 14 proceeds to step SP7, and based on the data stored in the memory 15, the average time between the J waves detected before the current J wave and the previous J wave from the current J wave. In the subsequent step SP8, it is determined whether or not the difference between the period and the average time is within a preset allowable range.

  Here, when this difference is outside the allowable range, this means that the interval between the current J wave and the immediately preceding J wave is irregular, and therefore the timing at which the toe of the right foot (left foot) takes off and the left foot There is a possibility of measuring an I-wave that does not appear due to the fact that the heel of the (right foot) touches down is close, or that tends to be in a low amplitude state even if it appears (FIG. 5 (B)). Means.

  Then, if a negative result is obtained in both step SP6 and step SP8, the one polarity initially set in step SP1 is based on the relationship between the road surface of the pedestrian and the charging sequence in the shoe sole (sole). This means that there is an extremely high possibility that a J wave appears in the other polarity having the opposite polarity.

  In this case, the processor 14 proceeds to step SP9, switches the setting of the polarity of the peak of the amplitude to be searched to the other polarity, and increments the step counter by “1” in the subsequent step SP10, Returning to SP2, the amplitude peak of the other polarity thus switched is searched from the walking waveform (for example, FIG. 5A) of the human body potential data supplied from the A / D converter 13.

  On the other hand, if an affirmative result is obtained in one or both of step SP6 and step SP8, this means that a J wave appears in one polarity initially set in step SP1, that is, the J wave is correctly generated. It means that there is a high possibility of being detected.

  In this case, the processor 14 increments the step count counter by “1” in step SP10 without switching the setting of the polarity of the amplitude peak to be searched, and then returns to step SP2 and thereafter performs A / D. From the walking waveform of human body potential data supplied from the converter 13 (for example, FIG. 5A), an amplitude peak of one of the initially set polarities is searched.

  In this manner, the processor 14 can execute the step count counting process.

  In the case of this embodiment, one of 0.2 [sec] or more and less than 0.3 [sec] is selected as the window time width set in step SP3. This is because it has been confirmed by experimental results that if it is selected for this time width, it can expect a certain level of accuracy, for example, it satisfies the standard (error ± 3 [%]) in the mechanical pedometer in Japanese Industrial Standards. It is.

  The experimental results are shown in FIGS. FIG. 11 shows the relationship between the time width of the window for the J wave, the number of times of erroneous counting as one step (error1 in the figure), and the number of times the actual number of steps could not be counted (error2 in the figure), FIG. 12 shows the relationship between the time width of the window for the I wave, the number of times of erroneous counting as one step (error1 in the figure), and the number of times the actual number of steps could not be counted (error2 in the figure). is there. By the way, the subjects in these experiments were 3 infants 4 to 5 years old (1 male, 2 females), 6 adults 20 to 40 years old (6 males), and 1 adult 60 years old. (Male).

  As is clear from this experimental result, the time width of the window is at least Japan if it is any of 0.2 [sec] or more and less than 0.3 [sec] for both the J wave and the I wave. It will meet the standards for mechanical pedometers in industrial standards.

(4) Operation and Effect In the above configuration, the pedometer 1 is an amplitude peak in one polarity with reference to a direct current component of electric field displacement (walking waveform) formed in the human body as the human body walks. In the predetermined time width centered on the searched amplitude peak, the maximum amplitude peak having the same polarity as the amplitude peak is set as the amplitude peak generated in the swing phase, and the step counter is set to 1. Only move up (count to one step).

  Therefore, this pedometer 1 corresponds to the case where one appears per step as described above in FIG. 4 and one foot is farthest from the road surface and the stance of the other foot is in a stable state. J-waves that have the least effect of interference between the component that appears as a free leg, that is, the component caused by the free leg and the component caused by the leg that is standing, can be detected as a reference for one step. Therefore, the number of steps can be accurately counted.

  In addition, the pedometer 1 does not immediately count the searched amplitude peak as one step, but within the predetermined time width centered on the amplitude peak, the maximum polarity having the same polarity as the amplitude peak. The amplitude peak is an amplitude peak that occurs during the swing phase.

  Therefore, the pedometer 1 can eliminate the amplitude peak caused by the potential change (noise) that does not correspond to walking, and therefore can more accurately count the number of steps.

  Further, the pedometer 1 automatically estimates whether the detection target J wave appears positively or negatively (step SP6 and step SP8), and is initialized as the polarity of the peak of the amplitude to be searched. When it is determined that no J wave appears in one polarity, the setting is switched to the other polarity.

  Therefore, the pedometer 1 can automatically correct the erroneous detection of the J wave. Therefore, the pedometer 1 can use the J wave as an index to determine the number of steps regardless of the road surface when measuring the number of steps and the charging series on the sole (sole). Can be counted.

  According to the above configuration, one that appears per step, and that appears when one leg is most separated from the road surface and the standing leg of the other leg is in a stable state, that is, a free leg. By making it possible to use the amplitude peak (J wave) that minimizes the influence of interference between the component caused by the feet that are standing and the component caused by the feet that are standing, as a reference for one step, The pedometer 1 that can accurately count the number of steps and thus can accurately detect walking can be realized.

(5) Other Embodiments In the above-described embodiment, the case where the parallel plate electrodes E1 and E2 are applied as the detection unit for detecting the electric field has been described. However, the present invention is not limited to this, For example, various other devices can be applied as long as they can detect an electric field, such as a micro dipole or an electro-optic crystal.

  In the above-described embodiment, the case where the BPF 12 that extracts the signal component in the frequency band of 20 [Hz] or less is provided in the front stage of the A / D converter 13, but the present invention is not limited thereto. Instead, it may be provided after the A / D converter 13.

  Further, in the above-described embodiment, whether the detection target J-wave appears positively or negatively (step SP6 and step SP8) is automatically set as the polarity of the peak of the amplitude to be searched. In the case where it is determined that the J wave does not appear in one polarity, the case where the setting is switched to the other polarity has been described. However, the present invention is not limited to this, and the sole (sole) is not limited thereto. The correspondence between the material of the road surface and the polarity of the appearance of the J wave is stored in the memory as a database, and the amplitude peak to be searched based on the material of the sole (sole) and the road surface that is input The polarity may be set.

  As a condition for switching the setting (hereinafter referred to as a setting switching condition), in the above-described embodiment, one of the polarity initially set between the current J wave and the immediately preceding J wave is The number of occurrences of the amplitude peak of the other polarity having the opposite polarity is 2 or more, and the average time between the J waves detected before the current J wave and the period from the current J wave to the immediately preceding J wave However, the present invention is not limited to this, and may satisfy any one of them.

  In the above-described embodiment, the setting is switched when the setting switching condition is not satisfied even once. However, the present invention is not limited to this, and the setting switching condition is not satisfied a predetermined number of times continuously. Alternatively, the setting may be switched.

  Furthermore, in the above-described embodiment, the case where the J wave is counted as one step has been described. However, in addition to or instead of this, the walking waveform between the J waves is identified by identifying the human body. The information may be extracted as information to be stored in the memory 15 as registered data. In this way, the registration data stored in the memory 15 can be applied to various authentication systems for authentication.

  The present invention can be used for a pedometer, an apparatus for registering a walking waveform as human body identification information, and the like.

It is a basic diagram which shows an electric field distribution pattern. It is a basic diagram which shows the dielectric constant in the biological tissue of the human body used for simulation. It is a basic diagram which shows the aspect of walking. It is a basic diagram which shows the electric potential change (walking waveform) of the electric field formed in a human body with one leg walking. It is a basic diagram which shows the electric potential change (walking waveform) of the electric field formed in a human body with walking on both legs. It is a basic diagram which shows the whole structure of the pedometer by this Embodiment. It is a basic diagram which shows an AA 'cross section. It is a basic diagram which shows an equivalent circuit. It is a basic diagram which shows the circuit structure of a main body. It is a flowchart which shows the step count processing procedure. It is a basic diagram which shows the relationship between the time width of the window with respect to J wave, and an error rate. It is a basic diagram which shows the time width of the window with respect to I wave, and the relationship of an error rate.

Explanation of symbols

  1 ... Pedometer, 2 ... Main unit, 3 ... Wristband, 11 ... Amplifier, 12 ... BPF, 13 ... A / D converter, 14 ... Processor, 15 ... Memory, E1 ... Human side Electrode, E2: outer electrode, PD: conductive pad.

Claims (5)

A first step of searching for an amplitude peak of one of the positive and negative polarities based on a direct current component from an electric field displacement in a frequency band of 20 [Hz] or less formed around the human body;
A second step of detecting a maximum amplitude peak having the same polarity as the amplitude peak within a predetermined time width centered on the searched amplitude peak as an amplitude peak generated in the swing phase; A walking waveform processing method characterized by comprising: The walking time processing method according to claim 1, wherein the time width is any one of 0.2 [sec] or more and less than 0.3 [sec]. Between the maximum amplitude peak detected as the current detection target and the maximum amplitude peak detected as the current detection target one time ago, the number of occurrences of an amplitude peak having a polarity opposite to the above polarity is 2 or more The walking waveform processing method according to claim 1, further comprising a third step of switching the setting of the polarity of the peak of the amplitude to be searched to the other polarity. Between the maximum amplitude peak detected as the current detection target and the maximum amplitude peak detected as the current detection target one time ago, the number of amplitude peak appearances having a polarity opposite to the above polarity is 2 or more And the average time between the peaks in each of the maximum amplitude peaks detected before the current detection target and the maximum amplitude peak detected as the current detection target are detected as the current detection target one time before. A third step of switching the setting of the polarity of the amplitude peak to be searched for to the other polarity when the difference from the period until the maximum amplitude peak is outside the predetermined allowable range; The walking waveform processing method according to claim 1. A detection unit for detecting an electric field;
An extraction unit for extracting a signal component in a frequency band of 20 [Hz] or less from the electric field detected by the detection unit;
A peak detection unit for detecting an amplitude peak generated during the swing phase from the signal component extracted by the extraction unit;
The peak detector is
The signal component is searched for an amplitude peak of one of the positive and negative polarities with reference to the DC component in the signal component, and within a predetermined time width centered on the searched amplitude peak. A walking waveform processing apparatus, wherein a maximum amplitude peak having the same polarity as the polarity of the amplitude peak is detected as an amplitude peak generated in the swing phase.