CN112244820B - Method for measuring running gait by triaxial accelerometer - Google Patents
Method for measuring running gait by triaxial accelerometer Download PDFInfo
- Publication number
- CN112244820B CN112244820B CN202011266523.2A CN202011266523A CN112244820B CN 112244820 B CN112244820 B CN 112244820B CN 202011266523 A CN202011266523 A CN 202011266523A CN 112244820 B CN112244820 B CN 112244820B
- Authority
- CN
- China
- Prior art keywords
- acceleration
- time
- frequency
- acc
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/112—Gait analysis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7253—Details of waveform analysis characterised by using transforms
- A61B5/7257—Details of waveform analysis characterised by using transforms using Fourier transforms
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
- G01P15/00—Measuring acceleration; Measuring deceleration; Measuring shock, i.e. sudden change of acceleration
- G01P15/18—Measuring acceleration; Measuring deceleration; Measuring shock, i.e. sudden change of acceleration in two or more dimensions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Surgery (AREA)
- Public Health (AREA)
- Signal Processing (AREA)
- Physiology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Veterinary Medicine (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Theoretical Computer Science (AREA)
- Psychiatry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Algebra (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses a method for measuring running gait by a triaxial accelerometer, which comprises the following steps: performing short-time domain spectrum analysis on the acquired real-time acceleration data to obtain a frequency band with the strongest energy entropy; constructing a filter based on the strongest frequency band of the energy entropy, and denoising the acceleration data; solving peak point data based on the denoised acceleration data; running gait is resolved based on the acceleration data and peak point data. According to the method for measuring running gait by the triaxial accelerometer, under the condition that the accelerometer is rigidly bound with a human body, a relatively simple gait estimation model is constructed, calculation of step frequency, step length, vertical amplitude and ground contact time length can be completed only by recording data of triaxial acceleration, the complex algorithm of the original multisensor and the training process of a large amount of data are simplified, resources occupied by an algorithm system are saved, hardware configuration is reduced, the set number of sensors of intelligent equipment is saved, enterprise cost is reduced, and great significance is achieved.
Description
Technical Field
The invention belongs to the technical field of intelligent motion detection equipment, and particularly relates to a method for measuring running gait by a triaxial accelerometer.
Background
With the rapid development of mobile devices and internet technologies, motion detection based on intelligent devices has gradually become a research hotspot in the fields of location services, artificial intelligence, and the like. The method for realizing motion state estimation at the present stage mainly adopts a multi-sensor fusion technology and a computer vision technology. The multi-sensor fusion technology collects signals of the advancing state of a user through an accelerometer, a gyroscope, a magnetometer and the like, and judges the state and gait parameters of pedestrians according to the signals and the change characteristics of the signals; the computer vision carries out model training through the labeling data of each gait, and uncalibrated data are substituted into the model to calculate gait parameters of the user. Both methods are modeling methods based on machine learning, and require a large amount of training data; meanwhile, the technology of gait recognition by using computer vision is established on video hardware equipment, and the motion scene of a user has great limitation.
Accordingly, the prior art is still further developed and improved.
Disclosure of Invention
In order to solve the above problems, a method for measuring running gait by using a three-axis accelerometer is proposed. The invention provides the following technical scheme:
a method of measuring running gait with a tri-axial accelerometer, comprising:
s100, performing short-time domain spectrum analysis on the acquired real-time acceleration data to obtain a frequency band with the strongest energy entropy;
s200, constructing a filter based on the strongest frequency band of the energy entropy, and denoising acceleration data;
s300, solving peak point data based on the denoised acceleration data;
s400, calculating running gait based on the acceleration data and the peak point data.
Further, the short-time-domain spectral analysis includes: after acquiring real-time acceleration data, carrying out Fourier transform on the acceleration data in the running state to realize frequency domain analysis, and determining the selection of a filter, passband cut-off frequency and stopband cut-off frequency based on the frequency band with the strongest energy entropy obtained by analysis.
Further, the process of denoising the acceleration data by the filter includes:
s210, constructing a filter to solve two coefficient vectors a= [ a (1), a (2) …, a (n) ] and b= [ b (1), b (2), …, b (n) ];
s220, performing filtering processing on the acceleration data:
where a (i) represents the i-th element of the a-vector, b (i) represents the i-th element of the b-vector, x (i) is the original acceleration of the input i-th epoch, and y (i) is the output corresponding to x (i).
Further, the process of solving the peak point data comprises:
s310, determining the sliding window length d based on the output frequency fre_acc and the running step frequency fre_step of the accelerometer:
s320, detecting and solving the maximum value of the ith acceleration value in the neighborhood region:
Acc_v(i)≥{Acc_v(k)|i-5＜k＜i+5} (3)；
wherein Acc_v (i) is the maximum value of the ith acceleration value in the neighborhood region;
if the window (3) is not satisfied, detecting the next window, otherwise, judging the next window:
Acc_v(i)≥Acc thr (4)
wherein Acc thr A threshold value that is a peak minimum value;
if the window (4) is met, then carrying out the next judgment, otherwise, detecting the next window:
time(i)-time(i-1)＞0.2s (5)
wherein time (i) refers to the time of the ith peak point.
Further, the running gait calculation process comprises step frequency detection, wherein the step frequency is the inverse of the time difference of two wave peaks:
fre_step(i)＝1/(time(i)-time(i-1)) (6)
where fre_step (i) refers to the step frequency at the i-th time.
Further, the running gait resolving process comprises step length estimation, regression fitting correction coefficients are made based on acceleration of chest and feet, a linear estimation model is built, and step length is resolved:
acc in norm Is the modulus of acceleration, acc x ，Acc y ，Acc z The acceleration values output by the accelerometer in the right direction, the front direction and the upper direction are respectively represented; (m, n) represents a step interval; Δa (i) represents the acceleration at the i-th step interval being extremely poor; step_len (i) is the step size of the i-th step interval.
Further, the running gait calculation process includes calculating a touchdown time length, collecting and constructing a mapping from the peak interval characteristic of the chest strap device to the touchdown time characteristic of the foot binding device based on the peak interval characteristic of the chest strap device and the touchdown time characteristic of the foot binding device:
Δt 2 ＝Z*Δt 1 +bias (8)
wherein t is 1 For a step interval calculated by chest acquisition acceleration and peak detection, t 2 For the stepping time obtained by the acceleration and feature extraction collected by the ankle, Z, bias is the slope and bias of the linear function, respectively.
Further, the running gait resolving process includes calculating vertical amplitude, constructing a trigonometric function curve based on an acceleration curve of a stepping interval, and solving a maximum amplitude of the trigonometric function curve:
approximating the acceleration waveform as a sin function:
Acc(t)＝A*sinωt (9)
where A is the magnitude of the trigonometric function, ω is the angular velocity, and t is the time between two peaks:
wherein A1 is the peak value of the trigonometric function, and A2 is the valley value of the trigonometric function;
and (3) performing double uncertain integration to obtain a displacement approximate function in the vertical direction:
the vertical amplitude ver_d (i) of the i-th step period is calculated:
where A is the magnitude of the trigonometric function and A (i) is the magnitude of the trigonometric function of the ith step period.
A computer readable storage medium having stored thereon a computer program which when executed by a processor implements a method of measuring running gait with a three-axis accelerometer.
An electronic terminal, comprising: a processor and a memory;
the memory is used for storing a computer program, and the processor is used for executing the computer program stored in the memory, so that the terminal executes a method for measuring running gait by the three-axis accelerometer.
The beneficial effects are that:
according to the method for measuring running gait by the triaxial accelerometer, under the condition that the accelerometer is rigidly bound with a human body, a relatively simple gait estimation model is constructed, calculation of step frequency, step length, vertical amplitude and ground contact time length can be completed only by recording data of triaxial acceleration, the complex algorithm of the original multisensor and the training process of a large amount of data are simplified, resources occupied by an algorithm system are saved, hardware configuration is reduced, the set number of sensors of intelligent equipment is saved, enterprise cost is reduced, and great significance is achieved.
Drawings
FIG. 1 is a flow chart of a method for measuring running gait with a three-axis accelerometer according to an embodiment of the invention;
FIG. 2 is a schematic diagram illustrating the interpretation of gait nouns in an embodiment of the invention;
FIG. 3 is a spectral analysis of acceleration data in an embodiment of the present invention;
FIG. 4 is a comparison of the frequency filtering before and after filtering in a specific embodiment of the invention;
FIG. 5 is a diagram of frequency filtering in a specific embodiment of the invention;
FIG. 6 is a graphical illustration of the filter processing acceleration in an embodiment of the present invention;
FIG. 7 is a schematic flow of peak detection in an embodiment of the invention;
FIG. 8 is a solution flow of peak detection in an embodiment of the invention;
FIG. 9 is a schematic diagram of step resolution in an embodiment of the invention;
FIG. 10 is an acceleration profile of two locations in an embodiment of the invention;
FIG. 11 is a graph showing the time relationship between the touchdown period and a step period in an embodiment of the present invention;
FIG. 12 is a graph of vertical amplitude resolution in an embodiment of the invention;
FIG. 13 illustrates the trigonometric calculation of vertical amplitude in an embodiment of the invention.
Detailed Description
In order to make the technical solution of the present invention better understood by those skilled in the art, the technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings, and based on the embodiments in the present application, other similar embodiments obtained by those skilled in the art without making creative efforts should fall within the scope of protection of the present application. In addition, directional words such as "upper", "lower", "left", "right", and the like, as used in the following embodiments are merely directions with reference to the drawings, and thus, the directional words used are intended to illustrate, not to limit, the invention.
As shown in fig. 1, a method for measuring running gait by a triaxial accelerometer includes:
s100, performing short-time domain spectrum analysis on the acquired real-time acceleration data to obtain a frequency band with the strongest energy entropy; the real-time acceleration data of the motion state of the human body is obtained under the condition that the triaxial accelerometer is rigidly bound with the human body, and the frequency spectrum analysis is carried out on the real-time acceleration data.
Further, the short-time-domain spectral analysis includes: after acquiring real-time acceleration data, carrying out Fourier transform on the acceleration data in the running state to realize frequency domain analysis, and determining the selection of a filter, passband cut-off frequency and stopband cut-off frequency based on the frequency band with the strongest energy entropy obtained by analysis. After the frequency spectrum analysis of the acceleration is completed, the energy value of the acceleration on each frequency band is obtained, the frequency band with the strongest energy entropy is intercepted, as shown in fig. 3, the walking state is strongest at 1-3 Hz, the running state is strongest at 2-5 Hz, and the data is used as the basis for constructing low-pass/band-pass filtering. For example, the band with the strongest energy entropy is (a, b), and aHz can be set to the passband cut-off frequency and bHz to the stopband cut-off frequency when the filter is constructed.
S200, constructing a filter based on the strongest frequency band of the energy entropy, and denoising acceleration data;
the purpose of constructing the filter is to extract the acceleration of a-bHz to achieve the denoising effect. Acceleration pairs before and after denoising are as shown in fig. 4.
Further, the process of denoising the acceleration data by the filter includes:
s210, constructing a filter to solve two coefficient vectors a= [ a (1), a (2) …, a (n) ] and b= [ b (1), b (2), …, b (n) ];
the frequency domain filtering can enable acceleration data to be smoother, and is beneficial to subsequent relevant detection and gait estimation work. As shown in fig. 5, since the acceleration data is time domain data, it is necessary to construct a filter to calculate two coefficient vectors a= [ a (1), a (2) …, a (n) ], and b= [ b (1), b (2), …, b (n) ].
S220, performing filtering processing on the acceleration data: the process of constructing the filter is an off-line process with the objective of obtaining a and b, so that the solution can be made using [ n, wn ] =button (wp, ws, ap, as) and [ b, a ] =button (n, wn) in Matlab. Wherein, button is Butterworth filter design function, and the wp of input, ws are passband cut-off frequency and stop band cut-off frequency after normalization respectively, ap, as are passband and stop band's maximum decay respectively. n and wn are the minimum order and the cut-off frequency of Butterworth respectively, and coefficient vectors b and a of the system function numerator and denominator polynomial of the Butterworth digital filter of the n order. After settlement b and a are obtained, the acceleration data can be subjected to filtering processing, as shown in fig. 6, and the formula of fig. 6 is as follows:
where a (i) represents the i-th element of the a-vector, b (i) represents the i-th element of the b-vector, x (i) is the original acceleration of the input i-th epoch, and y (i) is the output corresponding to x (i).
S300, solving peak point data based on the denoised acceleration data;
in the gait calculation process, the acceleration peak value is an important parameter, and the accuracy of detection of the acceleration peak value directly influences the calculation accuracy of the step frequency, the step width, the vertical amplitude and the touchdown time. After the filtering smoothing is completed, the acceleration data Acc with obvious peak characteristics and smooth waveform is obtained, wherein the Acc includes a time epoch time and a value acc_vertical (hereinafter abbreviated as acc_v) of acceleration in the vertical direction, and then the peak detection operation can be performed.
Further, the process of solving the peak point data comprises:
s310, determining the length d of the sliding window based on the output frequency fre_acc and the running step frequency fre_step of the accelerometer, wherein the peak detection is to find the maximum acceleration in a certain neighborhood region, and the length of the neighborhood, that is, the length d of the sliding window, needs to be determined before the detection is performed. Since the output frequency of the accelerometer is fre_acc:50Hz, running step frequency is about fre_step: 2-5 Hz, d is calculated by the following formula in order to accurately screen out the peak point of the acceleration:
taking fre_acc as 50Hz, d=10, i.e. the window size is 10 acceleration epochs.
S320, detecting and solving the maximum value of the ith acceleration value in the neighborhood region:
Acc_v(i)≥{Acc_v(k)|i-5＜k＜i+5} (3)；
where A is the magnitude of the trigonometric function and A (i) is the magnitude of the trigonometric function of the ith step period.
If the window (3) is not satisfied, detecting the next window, otherwise, judging the next window:
Acc_v(i)≥Acc thr (4)
wherein Acc thr A threshold value that is a peak minimum value;
if the window (4) is met, then carrying out the next judgment, otherwise, detecting the next window:
time(i)-time(i-1)＞0.2s (5)
wherein time (i) refers to the time of the ith peak point.
In the above process, equation (3) is the maximum value of the detection domain interval, equation (4) is to determine whether the value of the previous step is greater than the threshold value, equation (5) is whether the time interval between the current point and the previous peak point is greater than 0.2 seconds, and the schematic flow and the solution flow of the peak detection are shown in fig. 7 and 8, where T is used to store the time epoch of the peak point, and since the window length is 5 (the start interval), the start value of i is set to 5.
S400, calculating running gait based on the acceleration data and the peak point data.
As shown in fig. 2, the running gait includes ground contact time, stride frequency, stride length and vertical amplitude. The touchdown time refers to the duration of the heel strike to the toe off of a single foot; the step frequency refers to the number of steps in one second; stride refers to the distance in one stepping cycle; vertical amplitude refers to the height at which the center of gravity of the body moves vertically during running. Mapping the running process into acceleration characteristics can result in a stepping time and a stepping period.
Further, the running gait calculation process includes step frequency detection, and the acceleration of the user presents a distribution characteristic similar to sinewaves in the running process, so that the interval between two peaks can be regarded as a single step interval, and the step frequency of the time interval is equal to the reciprocal of the time difference between the two peaks:
fre_step(i)＝1/(time(i)-time(i-1)) (6)
where fre_step (i) refers to the step frequency at the i-th time.
Further, the running gait calculation process includes step size estimation, and according to the human morphology model, the fourth power root of the extremely poor acceleration in the vertical direction can approximately estimate the step size of the pedestrian, but the formula is only applicable to acceleration equipment with foot binding. Based on the acceleration of the chest and the foot, regression fit correction coefficients are made, a linear estimation model is built, as shown in fig. 9, and then the step length is solved:
acc in norm Is the modulus of acceleration, acc x ，Acc y ，Acc z The acceleration values output by the accelerometer in the right direction, the front direction and the upper direction are respectively represented; (m, n) represents a step interval; Δa (i) represents the acceleration at the i-th step interval being extremely poor; step_len (i) is the step size of the i-th step interval.
Further, the running gait calculation process includes calculating the touchdown time, and since the features acquired by the chest strap type device and the foot binding device are different, the features of the touchdown time cannot be acquired, and only the feature t of the peak interval can be acquired 1 The method comprises the steps of carrying out a first treatment on the surface of the The latter can obtain the touchdown time characteristic t more accurately 2 . According to this feature, construct t 1 To t 2 The map of (2) can be used to complete the measurement of touchdown time by chest strap data alone, as shown in fig. 10.
Acquiring and constructing a mapping from the peak interval characteristic of the chest strap device to the touchdown time characteristic of the foot binding device based on the peak interval characteristic of the chest strap device and the touchdown time characteristic of the foot binding device, as shown in fig. 11, t 1 And t 2 The linear relation exists between the two phases, namely an off-line phase and an on-line phase, can be set by utilizing a linear model to carry out regression fitting, the off-line phase works by collecting a large amount of sample data, taking a mapping function as a test sample, and obtaining the coefficient and the bias of the mapping function by a regression fitting methodSetting to obtain a mapping function f (t); the online stage is real-time settlement ground contact time length, and according to t 1 And f (t) calculating the touchdown time:
Δt 2 ＝Z*Δt 1 +bias (8)
wherein t is 1 For a step interval calculated by chest acquisition acceleration and peak detection, t 2 For the stepping time obtained by the acceleration and feature extraction collected by the ankle, Z, bias is the slope and bias of the linear function, respectively.
Further, the running gait calculation process includes calculating a vertical amplitude, where the vertical amplitude refers to a displacement of the center of gravity in the vertical direction, and an approximation value can be obtained by double integration of the acceleration in the vertical direction, but the settlement accuracy is greatly limited due to the limited sampling frequency, and the approximation diagram is shown in fig. 12.
According to the characteristic that the acceleration waveform is similar to the trigonometric function waveform, the acceleration curve of one stepping interval is approximated to be a trigonometric function curve, and the process of solving the vertical amplitude is changed into the process of solving the trigonometric function double integral. By means of the peak detection method, a trigonometric function curve is constructed based on the acceleration curve of a stepping interval at a time interval t which can reach a stepping period, a peak value A1 and a valley value A2 in the area, and the vertical amplitude, namely the maximum value of a displacement trigonometric function, namely the maximum amplitude of the trigonometric function is measured:
approximating the acceleration waveform as a sin function:
Acc(t)＝A*sinωwt (9)
where A is the magnitude of the trigonometric function, ω is the angular velocity, and t is the time between two peaks:
wherein A1 is the peak value of the trigonometric function, and A2 is the valley value of the trigonometric function;
and (3) performing double uncertain integration to obtain a displacement approximate function in the vertical direction:
the vertical amplitude ver_d (i) of the i-th step period is calculated:
where A is the magnitude of the trigonometric function and A (i) is the magnitude of the trigonometric function of the ith step period.
A computer readable storage medium having stored thereon a computer program which when executed by a processor implements a method of measuring running gait with a three-axis accelerometer.
An electronic terminal, comprising: a processor and a memory;
the memory is used for storing a computer program, and the processor is used for executing the computer program stored in the memory, so that the terminal executes a method for measuring running gait by the three-axis accelerometer.
The foregoing detailed description of the invention has been presented for purposes of illustration and description, but is not intended to limit the scope of the invention, i.e., the invention is not limited to the details shown and described.
Claims (3)
1. A method of measuring running gait with a tri-axial accelerometer, comprising:
s100, performing short-time domain spectrum analysis on the acquired real-time acceleration data to obtain a frequency band with the strongest energy entropy;
s200, constructing a filter based on the strongest frequency band of the energy entropy, and denoising acceleration data;
s300, solving peak point data based on the denoised acceleration data;
s400, calculating running gait based on acceleration data and peak point data;
the short-time domain spectral analysis includes: after acquiring real-time acceleration data, carrying out Fourier transform on the acceleration data in the running state to realize frequency domain analysis, and determining the selection of a filter, passband cut-off frequency and stopband cut-off frequency based on the frequency band with the strongest energy entropy obtained by analysis;
the process of denoising the acceleration data by the filter comprises the following steps:
s210, constructing a filter to solve two coefficient vectors a= [ a (1), a (2) …, a (n) ] and b= [ b (1), b (2), …, b (n) ];
s220, performing filtering processing on the acceleration data:
wherein a (i) represents the ith element of the a vector, b (i) represents the ith element of the b vector, x (i) is the original acceleration of the input ith epoch, and y (i) is the output corresponding to x (i);
the process for solving the peak point data comprises the following steps:
s310, determining the sliding window length d based on the output frequency fre_acc and the running step frequency fre_step of the accelerometer:
s320, detecting and solving the maximum value of the ith acceleration value in the neighborhood region:
Acc_v(i)≥{Acc_v(k)|i-5＜k＜i+5} (3)；
wherein acc_v (i) is the maximum value of the ith acceleration value in the neighborhood region, if (3) is not satisfied, then detecting the next window, otherwise, performing the next step of judgment:
Acc_v(i)≥Acc thr (4)
wherein Acc thr A threshold value that is a peak minimum value;
if the window (4) is met, then carrying out the next judgment, otherwise, detecting the next window:
time(i)-time(i-1)＞0.2s (5)
wherein time (i) refers to the time of the ith peak point;
the running gait resolving process comprises step frequency detection, wherein the step frequency is the reciprocal of the time difference of two wave peaks:
fre_step(i)＝1/(time(i)-time(i-1)) (6)
wherein fre_step (i) refers to the step frequency at the i-th moment; when the acceleration data is subjected to the filtering process: the process of constructing the filter is an off-line process, the purpose is to obtain a and b, and the solution is performed by using [ n, wn ] = button (wp, ws, ap, as) and [ b, a ] = button (n, wn) in Matlab; wherein, button is Butterworth filter design function, the wp and ws of input are passband cut-off frequency and stop band cut-off frequency after normalizing, ap, as are the maximum attenuation of passband and stop band respectively; n and wn are respectively the minimum order and the cut-off frequency of Butterworth, and the coefficient vectors b and a of the system function numerator and denominator polynomial of the Butterworth digital filter of the n order; after b and a are obtained through calculation, the acceleration data can be subjected to filtering processing;
the running gait resolving process comprises step length estimation, regression fitting correction coefficients are made based on acceleration of chest and feet, a linear estimation model is built, and step length is resolved:
acc in norm Is the modulus of acceleration, acc x ，Acc y ，Acc z The acceleration values output by the accelerometer in the right direction, the front direction and the upper direction are respectively represented; (m, n) represents a step interval; Δa (i) represents the acceleration at the i-th step interval being extremely poor; step_len (i) is the step size of the i-th step interval;
the running gait resolving process comprises the steps of calculating the touchdown time, collecting and constructing a mapping from the peak interval characteristic of chest strap equipment to the touchdown time characteristic of foot binding equipment based on the peak interval characteristic of the chest strap equipment and the touchdown time characteristic of foot binding equipment, specifically, collecting and constructing a mapping from the peak interval characteristic of the chest strap equipment to the touchdown time characteristic of foot binding equipment based on the peak interval characteristic of the chest strap equipment and the touchdown time characteristic of foot binding equipment, wherein an approximately linear relation exists between t1 and t2, regression fitting is carried out by utilizing a linear model, two stages of offline and online are set, the offline stage is operated by collecting a large amount of sample standard data, serving as a test sample for a mapping function, obtaining the coefficient and bias of the mapping function through a regression fitting method, and obtaining the mapping function f (t); the online stage is to settle the touchdown time in real time, and the touchdown time is solved according to t1 and f (t):
Δt 2 ＝Z*Δt 1 +bias (8)
wherein t is 1 For a step interval calculated by chest acquisition acceleration and peak detection, t 2 For the stepping time obtained by the acceleration and feature extraction collected by the ankle, Z, bias is the slope and bias of the linear function, respectively;
the running gait resolving process comprises the steps of calculating vertical amplitude, constructing a trigonometric function curve based on an acceleration curve of a stepping interval, and solving the maximum amplitude of the trigonometric function curve:
approximating the acceleration waveform as a sin function:
Acc(t)＝A*sinωt (9)
where A is the magnitude of the trigonometric function, ω is the angular velocity, and t is the time between two peaks:
wherein A1 is the peak value of the trigonometric function, and A2 is the valley value of the trigonometric function; and (3) performing double uncertain integration to obtain a displacement approximate function in the vertical direction:
the vertical amplitude ver_d (i) of the i-th step period is calculated:
where A is the magnitude of the trigonometric function and A (i) is the magnitude of the trigonometric function of the ith step period.
2. A computer-readable storage medium having stored thereon a computer program, characterized by: which when executed by a processor implements the method as claimed in claim 1.
3. An electronic terminal, comprising: a processor and a memory;
the memory is configured to store a computer program, and the processor is configured to execute the computer program stored in the memory, so as to cause the terminal to perform the method as set forth in claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011266523.2A CN112244820B (en) | 2020-11-13 | 2020-11-13 | Method for measuring running gait by triaxial accelerometer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011266523.2A CN112244820B (en) | 2020-11-13 | 2020-11-13 | Method for measuring running gait by triaxial accelerometer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112244820A CN112244820A (en) | 2021-01-22 |
| CN112244820B true CN112244820B (en) | 2023-05-26 |
Family
ID=74266762
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011266523.2A Active CN112244820B (en) | 2020-11-13 | 2020-11-13 | Method for measuring running gait by triaxial accelerometer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112244820B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113659962A (en) * | 2021-08-03 | 2021-11-16 | 青岛迈金智能科技有限公司 | Disc claw stepping frequency meter and parameter optimization method for same |
| CN113465733B (en) * | 2021-08-13 | 2022-04-05 | 重庆大学 | Vibration table structure displacement response prediction method and device based on EEMD-DNN |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109195104A (en) * | 2018-08-27 | 2019-01-11 | 上海市计量测试技术研究院 | A kind of indoor orientation method combined based on support vector regression and Kalman filtering |
| WO2019043601A1 (en) * | 2017-08-29 | 2019-03-07 | Myotest Sa | A method and device for retrieving biomechanical parameters of a stride |
| CN109827577A (en) * | 2019-03-26 | 2019-05-31 | 电子科技大学 | High-precision inertial navigation location algorithm based on motion state detection |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6018705A (en) * | 1997-10-02 | 2000-01-25 | Personal Electronic Devices, Inc. | Measuring foot contact time and foot loft time of a person in locomotion |
| US6826509B2 (en) * | 2000-10-11 | 2004-11-30 | Riddell, Inc. | System and method for measuring the linear and rotational acceleration of a body part |
| US10285625B2 (en) * | 2013-10-07 | 2019-05-14 | Wahoo Fitness Llc | Activity monitoring computing device and system |
| US9811720B2 (en) * | 2013-10-22 | 2017-11-07 | Bae Systems Information And Electronic Systems Integration Inc. | Mobile device based gait biometrics |
| CN103983273B (en) * | 2014-04-29 | 2017-06-06 | 华南理工大学 | A kind of real-time step-size estimation method based on acceleration transducer |
| CN106289309B (en) * | 2016-10-26 | 2019-08-16 | 深圳大学 | Step-recording method and device based on 3-axis acceleration sensor |
| CN109115207B (en) * | 2017-06-23 | 2021-06-04 | 北京方位捷讯科技有限公司 | Pedestrian walking track detection method, device and system |
| CN108444473B (en) * | 2018-03-20 | 2020-08-28 | 南京华苏科技有限公司 | Indoor pedestrian trajectory positioning method |
| CN109498027A (en) * | 2018-12-19 | 2019-03-22 | 南京茂森电子技术有限公司 | A kind of list accelerometer body gait detection system and method |
| KR102304300B1 (en) * | 2019-05-08 | 2021-09-23 | 주식회사 비플렉스 | A method and apparatus for detecting walking factor with portion acceleration sensor |
| CN111536989B (en) * | 2020-05-26 | 2021-08-24 | 江南大学 | Adaptive time window step counting method based on peak detection |
| CN111854738A (en) * | 2020-06-22 | 2020-10-30 | 厦门亿联网络技术股份有限公司 | Human motion state detection method and device based on three-axis accelerometer |
-
2020
- 2020-11-13 CN CN202011266523.2A patent/CN112244820B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019043601A1 (en) * | 2017-08-29 | 2019-03-07 | Myotest Sa | A method and device for retrieving biomechanical parameters of a stride |
| CN109195104A (en) * | 2018-08-27 | 2019-01-11 | 上海市计量测试技术研究院 | A kind of indoor orientation method combined based on support vector regression and Kalman filtering |
| CN109827577A (en) * | 2019-03-26 | 2019-05-31 | 电子科技大学 | High-precision inertial navigation location algorithm based on motion state detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112244820A (en) | 2021-01-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112244820B (en) | Method for measuring running gait by triaxial accelerometer | |
| CN104866099B (en) | The error compensating method of smart machine gesture identification precision is improved based on motion sensor | |
| CN103679171B (en) | A gait feature extraction method based on human body gravity center track analysis | |
| CN107220617A (en) | Human body attitude identifying system and method | |
| CN103099611B (en) | Interference suppression system for sphygmomanometer measurement and interference suppression method thereof | |
| CN104408718B (en) | A kind of gait data processing method based on Binocular vision photogrammetry | |
| Susi et al. | Accelerometer signal features and classification algorithms for positioning applications | |
| CN105180959B (en) | A kind of anti-interference step-recording method suitable for wrist pedometer | |
| WO2016180653A1 (en) | Pedometer and method for analyzing motion data | |
| EP3090685A1 (en) | Pedometer and method for analyzing motion data | |
| Shao et al. | DePedo: Anti periodic negative-step movement pedometer with deep convolutional neural networks | |
| CN114469073B (en) | Gait analysis and abnormality detection method based on wearable sensor | |
| WO2023108315A1 (en) | Gait analysis and anomaly detection method based on wearable step sensor | |
| US20060268986A1 (en) | Process for estimating the motion phase of an object | |
| CN106214143A (en) | Transient motion interference identification method and heart rate confidence calculations method | |
| CN116226727A (en) | Motion recognition system based on AI | |
| CN114440883B (en) | Pedestrian positioning method based on foot and leg micro-inertia measurement unit | |
| CN116206358A (en) | Lower limb exoskeleton movement mode prediction method and system based on VIO system | |
| Briassouli et al. | Continuous wavelet transform for time-varying motion extraction | |
| CN105232053A (en) | Human ankle joint plantar flexion phase detection system and method | |
| CN114533010A (en) | Heart rate detection method and device | |
| CN113128474A (en) | Structural modal identification method based on computer vision and variational modal decomposition | |
| CN113790722B (en) | Pedestrian step length modeling method based on inertial data time-frequency domain feature extraction | |
| CN112560002A (en) | Gait behavior-based identity authentication method, device, equipment and storage medium | |
| Molton | Computer vision as an aid for the visually impaired |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 266000 Room 302, building a 3, 328 Chengkang Road, Xiazhuang street, Chengyang District, Qingdao City, Shandong Province Applicant after: Qingdao maijin Intelligent Technology Co.,Ltd. Address before: 266000 room 207-1, building 2, 130 jiushui East Road, Fushan road street, Licang District, Qingdao City, Shandong Province Applicant before: QINGDAO MAGENE INTELLIGENCE TECHNOLOGY Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |