CN105268171B - gait monitoring method, device and wearable device - Google Patents
gait monitoring method, device and wearable device Download PDFInfo
- Publication number
- CN105268171B CN105268171B CN201510561183.9A CN201510561183A CN105268171B CN 105268171 B CN105268171 B CN 105268171B CN 201510561183 A CN201510561183 A CN 201510561183A CN 105268171 B CN105268171 B CN 105268171B
- Authority
- CN
- China
- Prior art keywords
- data
- user
- gait
- characteristic
- training
- 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
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
A kind of gait monitoring method of the application offer, device and wearable device, this method include：Obtain corresponding multiple data segments during user's sole during running contacts to earth；The characteristic for carrying out gait analysis is determined from the multiple data segment；The characteristic is input in the mathematical model trained, gait monitoring result of user during running is obtained.It can allow the user to correctly recognize the gait of oneself in the case where no professional person instructs in technical scheme of the present invention, and adjustment appropriate is made to the gait of user according to gait monitoring result, improve the running level of user and avoids sport injury.
Description
Technical field
This application involves wearable device technical field more particularly to a kind of gait monitoring method, device and wearable set
It is standby.
Background technology
With the development and the improvement of people's living standards of society, body-building is increasingly becoming the important need of people.In China,
Nationwide fitness programs have been developed as a mass movement.And run as a kind of simple and practicable body-building, increasingly by
People's dotes on.But the jog mode of mistake may cause sport injury to human body to some extent, therefore correctly into
Row road-work is extremely important to running crowd.And most of crowd for carrying out road-work cannot get the finger of professional person
It leads, running posture can only be adjusted with self-observation by learning by oneself.
Invention content
In view of this, the application provides a kind of new technical solution, the gait prison during user can be made to pass through running
It surveys result and adjustment appropriate is made to gait, improve the running level of user and avoid sport injury.
To achieve the above object, it is as follows to provide technical solution by the application：
According to the first aspect of the application, it is proposed that a kind of gait monitoring method, including：
Obtain corresponding multiple data segments during sole of user during running contacts to earth；
The characteristic for carrying out gait analysis is determined from the multiple data segment；
The characteristic is input in the mathematical model trained, gait of user during running is obtained
Monitoring result.
According to the second aspect of the application, it is proposed that a kind of gait monitoring device, including：
First acquisition module, for obtaining corresponding multiple data during sole of user during running contacts to earth
Section；
First determining module, the multiple data segment for being got from first acquisition module are determined for carrying out
The characteristic of gait analysis；
First computing module, the characteristic for determining first determining module are input to the number trained
It learns in model, obtains gait monitoring result of user during running.
According to the third aspect of the application, it is proposed that a kind of wearable device, the wearable device include：
Processor；Memory for storing the processor-executable instruction；
Wherein, the processor, is configured as：
Obtain corresponding multiple data segments during sole of user during running contacts to earth；
The characteristic for carrying out gait analysis is determined from the multiple data segment；
The characteristic is input in the mathematical model trained, gait of user during running is obtained
Monitoring result.
By above technical scheme as it can be seen that the application is realized is monitored to gait of user during running, to
It can allow the user to correctly recognize the gait of oneself in the case where no professional person instructs, and according to gait monitoring result
Adjustment appropriate is made to the gait of user, improve the running level of user and avoids sport injury.
Description of the drawings
Figure 1A shows the flow diagram of the gait monitoring method of an exemplary embodiment one according to the present invention；
Figure 1B show data that the front foot of Figure 1A illustrated embodiments lands time domain schematic diagram；
Fig. 1 C show data that the metapedes of Figure 1A illustrated embodiments lands time domain schematic diagram；
Fig. 2A shows the flow diagram of the gait monitoring method of an exemplary embodiment two according to the present invention；
Fig. 2 B show data segment that the front foot that Fig. 2A illustrated embodiments obtain lands time domain schematic diagram；
Fig. 2 C show data segment that the metapedes that Fig. 2A illustrated embodiments obtain lands time domain schematic diagram；
Fig. 3 A show the flow diagram of the gait monitoring method of an exemplary embodiment three according to the present invention；
Fig. 3 B show data that the front foot of Fig. 3 A illustrated embodiments lands frequency domain schematic diagram；
Fig. 3 C show data that the metapedes of Fig. 3 A illustrated embodiments lands frequency domain schematic diagram；
Fig. 4 shows the flow diagram of the gait monitoring method of an exemplary embodiment four according to the present invention；
Fig. 5 shows the flow diagram of the gait monitoring method of an exemplary embodiment five according to the present invention；
Fig. 6 A show the flow diagram of the gait monitoring method of an exemplary embodiment six according to the present invention；
Fig. 6 B show the model schematic based on support vector machines of Fig. 6 A illustrated embodiments；
Fig. 7 A show the flow diagram of the gait monitoring method of an exemplary embodiment seven according to the present invention；
Fig. 7 B show the model schematic based on decision tree of Fig. 7 A illustrated embodiments；
Fig. 8 shows the structural schematic diagram of the wearable device of an exemplary embodiment according to the present invention；
Fig. 9 shows the structural schematic diagram of the gait monitoring device of an exemplary embodiment according to the present invention；
Figure 10 shows the structural schematic diagram of gait monitoring device in accordance with an alternative illustrative embodiment of the present invention；
Figure 11 shows the structural schematic diagram of gait monitoring device in accordance with a further exemplary embodiment of the present invention.
Specific implementation mode
Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to
When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment
Described in embodiment do not represent all embodiments consistent with the application.On the contrary, they be only with it is such as appended
The example of consistent device and method of some aspects be described in detail in claims, the application.
It is the purpose only merely for description specific embodiment in term used in this application, is not intended to be limiting the application.
It is also intended to including majority in the application and "an" of singulative used in the attached claims, " described " and "the"
Form, unless context clearly shows that other meanings.It is also understood that term "and/or" used herein refers to and wraps
Containing one or more associated list items purposes, any or all may be combined.
It will be appreciated that though various information, but this may be described using term first, second, third, etc. in the application
A little information should not necessarily be limited by these terms.These terms are only used for same type of information being distinguished from each other out.For example, not departing from
In the case of the application range, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as
One information.Depending on context, word as used in this " if " can be construed to " ... when " or " when ...
When " or " in response to determination ".
For general running personage comes, the main contents of jog mode include cadence, with speed and gait.Cadence indicates
User's quantity per minute taken a step indicates the running time needed for every kilometer of user with speed, and gait expression user was running
The mode that sole in journey lands, during actually running, the difference for the sequence that landed according to each position of sole, gait is divided into
Forward roll, the rear foot heelstrike land with full sole.According to the actual needs, forward roll and full sole can be landed
Be collectively referred to as front foot to land, and the rear foot be heelstrike known as to metapedes and is landed, or forward roll is known as front foot and is landed, and will after
It heelstrike lands with full sole and is collectively referred to as metapedes and lands.The above-mentioned different classifications method of gait, can pass through the step of the application
State monitoring method is realized.The application is by being monitored gait of user during running, so as to allow users to
It can also correctly recognize the gait of oneself in the case where no professional person instructs, and the gait of user is done according to gait monitoring result
Go out adjustment appropriate, improve the running level of user and avoids sport injury.
For the application is further described, the following example is provided：
Figure 1A shows the flow diagram of the gait monitoring method of an exemplary embodiment one according to the present invention, Figure 1B
Data that the front foot of Figure 1A illustrated embodiments lands are shown in the schematic diagram of time domain, Fig. 1 C show Figure 1A illustrated embodiments
Schematic diagram of the data that metapedes lands in time domain；The present embodiment can by mounted on running equipment on gait monitoring device come
It realizes, wherein running equipment can include but is not limited to running shoes, insole and socks etc..The gait monitoring device is in running equipment
On installation site include but not limited to running shoes sole, vamp and heel position, the arch of foot position of insole and the socks of socks
Cylinder etc., as shown in Figure 1A, gait monitoring method includes the following steps：
Step 101, corresponding multiple data segments during sole of user during running contacts to earth are obtained；
Step 102, the characteristic for carrying out gait analysis is determined from multiple data segments；
Step 103, characteristic is input in the mathematical model trained, obtains gait of user during running
Monitoring result.
In a step 101, sensing data of user during running, the sensor number can be acquired by sensor
According to can be by including but not limited to be collected by sensors such as acceleration transducer, gyroscope and magnetometers.From biography
Identified in sensor data sole contact to earth during corresponding data segment.As illustrated in figures ib and 1 c, it is that acceleration transducer exists
The sensing data of collected one dimension wherein of marking time during user's running, wherein Figure 1B shows user
Front foot land brief acceleration sensor acquisition data, Fig. 1 C show user's metapedes land brief acceleration sensor acquisition number
According to, in the data of this sensor, front foot land with metapedes land contact to earth during be corresponding with higher wave crest, respective
Wave crest both sides have the trough of different shape, the data characteristics of the trough by detecting wave crest and wave crest both sides, so as to
It determines user's corresponding data segment during the sole of each step contacts to earth, and then use is able to detect that during entire running
The corresponding multiple data segments of step number and the step number that family is run.When using multiple dimension datas of the same sensor, then
Each step of user during running can all correspond to the corresponding data segment of multiple dimensions during contacting to earth, further, when
When having multiple sensors to acquire, each sensor in multiple sensors can collect the corresponding data of respective different dimensions
Section, the application to simplify the explanation, are only illustrated with the data of a dimension of acceleration transducer；This field skill
Art personnel are it is understood that the processing mode of the data of the collected different dimensions of different sensors may refer to the application's
Associated description, and the description of the data of a dimension of the acceleration transducer in the application cannot form the limit to the application
System.
In a step 102, in one embodiment, multiple data segments can be determined in time domain for carrying out gait analysis
Characteristic；In another embodiment, the characteristic for carrying out gait analysis can be determined in frequency domain to multiple data segments,
Detailed description refers to following Fig. 3 A and embodiment illustrated in fig. 4, is not described in detail first herein.
In step 103, in one embodiment, gait monitoring result can be considered as to a binary classification problems, i.e. root
Judge that user is that front foot lands or metapedes lands according to the characteristic extracted.In one embodiment, the mathematical modulo trained
Type can be the mathematical model based on biomechanical analysis, and land the mechanics difference to land with metapedes according to front foot, formulates respectively
Front foot lands the feature templates to land with metapedes, and by the method for Dynamic Programming by the data characteristics of user and given feature
Template is matched, and is that front foot lands or metapedes lands according to matching distance judgement, wherein each dimensional feature in feature templates
Parameter value can be by testing iteration adjustment.In another embodiment, the mathematical model trained can be machine learning mathematics
Model is carried out the parameter training of mathematical model by front foot data known to a large amount of gaits or metapedes data, and training is obtained
Parameter model be used for user data Gait Recognition.In one embodiment, machine learning mathematical model include but not limited to away from
From similarity model, perceptron model, supporting vector machine model, decision-tree model, etc..
Seen from the above description, the realization of 101- steps 103 was running to user to the embodiment of the present invention through the above steps
Gait in journey is monitored, so as to allow users to correctly recognize oneself in the case where no professional person instructs
Gait, and adjustment appropriate is made to the gait of user according to gait monitoring result, improve the running level of user and avoids transporting
Dynamic injury.
Fig. 2A shows the flow diagram of the gait monitoring method of an exemplary embodiment two according to the present invention, Fig. 2 B
Show that the front foot that Fig. 2A illustrated embodiments obtain lands in the schematic diagram of the data segment of time domain, Fig. 2 C show real shown in Fig. 2A
Apply data segment that the metapedes that example obtains lands time domain schematic diagram；As shown in Figure 2 A, include the following steps：
Step 201, determine whether user is in running state；
Step 202, if user is in running state, sensing data of user during running is obtained；
Step 203, signal filtering is carried out to sensing data；
Step 204, each step sole of the user that the sensing data after detection filter detects during running touches
Corresponding wave crest and trough during ground；
Step 205, determine that sole of user during running contacts to earth process according to the corresponding wave crest of each step and trough
In corresponding multiple data segments.
In step 201, it can determine whether user is in running state by the step number in the unit interval, for example,
When detecting that step number of the user within the unit interval reached setting step number, it is determined that user is in running state, wherein setting
Step number can be determined by actual tests.
In step 203 to step 205, it includes but not limited to low-pass filtering, card to carry out signal filtering to sensing data
The filtering of Germania rate, medium filtering etc..It can determine time point of the user when contacting to earth, be intercepted according to time point when contacting to earth tactile
Data segment before and after ground.The front and back data segment that contacts to earth may refer to Fig. 2 B and Fig. 2 C, so as to by except two vertical lines
Data during non-during running is contacted to earth filter out.
In the present embodiment, contacted to earth by obtaining sole of user during running after being pre-processed sensing data
Corresponding multiple data segments in the process reduce calculation amount, improve meter to make unnecessary data be not involved in the processing in later stage
Efficiency is calculated, while improving the precision of gait analysis.
Fig. 3 A show the flow diagram of the gait monitoring method of an exemplary embodiment three according to the present invention, Fig. 3 B
Data that the front foot of Fig. 3 A illustrated embodiments lands are shown in the schematic diagram of frequency domain, Fig. 3 C show Fig. 3 A illustrated embodiments
Schematic diagram of the data that metapedes lands in frequency domain；The present embodiment in frequency domain by the front and back data segment that contacts to earth to obtain characteristic
For illustrate, as shown in Figure 3A, include the following steps：
Step 301, corresponding multiple data segments during sole of user during running contacts to earth are obtained；
Step 302, FFT transform is carried out to multiple data segments, obtains multiple data segments in the corresponding energy spectrum of frequency domain；
Step 303, the characteristic for indicating Energy distribution is determined on energy spectrum, wherein the feature of Energy distribution
Data are for carrying out gait analysis；
Step 304, by for indicating that the characteristic of Energy distribution is input in the mathematical model trained, user is obtained
Gait monitoring result during running.
The description of step 301 and step 304 may refer to the description of related embodiment, and this will not be detailed here.
In step 302 and step 303, data segment when by that will contact to earth carries out FFT transform, obtains sensing data section
In the energy spectrum information of frequency domain, as shown in Fig. 3 B and Fig. 3 C, embody that front foot lands and metapedes lands corresponding contacted to earth
The Energy distribution of journey on a different frequency, for example, can be contacted to earth by summing to the energy spectrum during contacting to earth
Gross energy in journey, using gross energy as the characteristic of Energy distribution, for another example to setting frequency range (for example, 1-25Hz or
Between 25Hz-50Hz) energy spectrum be identified, using the energy values of given frequency range as the characteristic of Energy distribution.
In the present embodiment, frequency domain is transformed by corresponding sensing data section during contacting to earth and carries out data characteristics
Analysis, different energy Spectral structures have been corresponded to since front foot lands to land from metapedes, are used for by determining on energy spectrum
Indicate that the characteristic of Energy distribution can greatly improve the accuracy of Gait Recognition.
Fig. 4 shows the flow diagram of the gait monitoring method of an exemplary embodiment four according to the present invention；This reality
Example is applied by by contacting to earth front and back data segment for time domain obtains characteristic and Fig. 2 B and Fig. 2 C being combined to carry out exemplary theory
It is bright, as shown in figure 4, including the following steps：
Step 401, corresponding multiple data segments during sole of user during running contacts to earth are obtained；
Step 402, multiple data segments are determined with corresponding mean value, wave crest within each data segment corresponding period
Feature, trough feature, peak valley concussion value and peak valley spacing, wherein mean value, wave crest feature, trough feature, peak valley concussion value with
And peak valley spacing is as the data characteristics for carrying out gait analysis；
Step 403, mean value, wave crest feature, trough feature, peak valley concussion value and peak valley spacing are input to and have been trained
In mathematical model, gait monitoring result of user during running is obtained.
The description of step 401 and step 403 may refer to the description of related embodiment, and this will not be detailed here.
As shown in fig. 2 b and fig. 2 c, exemplary theory is carried out with the data instance of the collected dimension of acceleration transducer
It is bright, during the mean value of the data segment during contacting to earth indicates that acceleration mean value during contacting to earth, wave crest expression are contacted to earth
Peak-peak, trough indicate during contacting to earth minimum valley value (such as Fig. 2 B and two vertical lines shown in fig. 2 C and wave crest it
Between minimum valley value), peak valley concussion value is the difference of maximum crest value and minimum valley value, peak valley spacing be maximum wave crest with
At a distance of the number number of sampled point between minimum trough.
It is special by the mean value in corresponding sensing data section during contacting to earth, wave crest feature, trough in the present embodiment
Sign, peak valley concussion value and peak valley spacing since front foot lands with metapedes as the data characteristics for carrying out gait analysis
Ground has corresponded to different mean values, wave crest feature, trough feature, peak valley concussion value and peak valley spacing, and the data of these time domains are special
Sign can be separately as the input for the mathematical model trained, in addition it is also possible to the data characteristics of frequency domain together as having instructed
The input of experienced mathematical model, to improve the precision of result of calculation.Compared with the data characteristics of frequency domain, due to the data of time domain
Feature need not carry out FFT transform, therefore the advantage that the data characteristics of time domain has computation complexity low.
Fig. 5 shows the flow diagram of the gait monitoring method of an exemplary embodiment five according to the present invention；This reality
Apply example and illustrated so that the mathematical model trained is Distance conformability degree model as an example, adjust the distance similarity model into
When row parameter training, it is thus necessary to determine that front foot data and metapedes data included in training data, wherein training data is for instructing
Practice Distance conformability degree model, then calculate separately front foot data in Distance conformability degree model corresponding first distributed constant and after
Sufficient data corresponding second distributed constant in Distance conformability degree model, the first distributed constant and the second distributed constant can determine
Go out Distance conformability degree model.
As shown in figure 5, including the following steps：
Step 501, corresponding multiple data segments during sole of user during running contacts to earth are obtained；
Step 502, the characteristic for carrying out gait analysis is determined from multiple data segments；
Step 503, by characteristic respectively with the first distributed constant and the second distributed constant in Distance conformability degree model
Similarity calculation is carried out, the first similarity and the second similarity are obtained；
Step 504, gait monitoring result of user during running is determined according to the first similarity and the second similarity.
The description of step 501 and step 502 may refer to the description of above-mentioned related embodiment, and this will not be detailed here.
In step 503, in one embodiment, can by include but not limited to by it is in the prior art it is European away from
Obtain the first similarity in the application and the second similarity from, mahalanobis distance etc., the first similarity and the second similarity with
Either mahalanobis distance is inversely proportional that is, Euclidean distance or mahalanobis distance are smaller the Euclidean distance being calculated, and similarity corresponds to
Value it is bigger.How the present embodiment is to by Euclidean distance, mahalanobis distance to obtain the first similarity in the application and the second phase
It is described like without expansion, the prior art can be referred to.
In step 504, the first similarity can be compared with the second similarity, by the first similarity and the second phase
Gait like corresponding to the higher value in degree as gait monitoring result, for example, if the first similarity be more than the second similarity,
Indicate that the corresponding Euclidean distance of the first similarity or mahalanobis distance are smaller, then the corresponding gait of the data segment is the first similarity
Corresponding front foot lands, if the first similarity is less than the second similarity, indicate the corresponding Euclidean distance of the second similarity or
Mahalanobis distance is smaller, then the corresponding gait of the data segment is that the corresponding metapedes of the second similarity lands.
In the present embodiment, by by characteristic respectively in Distance conformability degree model the first distributed constant and second point
Cloth parameter carry out similarity calculation, obtain the first similarity and the second similarity, by the first similarity and the second similarity come
Gait monitoring result of user during running is determined, to convert gait monitoring problem for binary classification problems, also
The accuracy of gait monitoring result can be continuously improved by using the statistical model parameter that mass data obtains.
Fig. 6 A show the flow diagram of the gait monitoring method of an exemplary embodiment six according to the present invention, Fig. 6 B
Show the model schematic based on support vector machines of Fig. 6 A illustrated embodiments；The present embodiment is with the mathematical model trained
It is illustrated for supporting vector machine model, as shown in Figure 6B, when carrying out parameter training to supporting vector machine model,
Determine front foot data included in training data (in Fig. 6 B shown in "×") and metapedes data (in Fig. 6 B shown in " Ο "),
In, training data is used for Training Support Vector Machines model；Training one according to front foot data and metapedes data makes front foot land
And the metapedes maximum Optimal Separating Hyperplane function of spacing distance between corresponding two classifications that lands passes through instruction as shown in Figure 6B
Practice data and obtained two supporting vector hyperplane and an Optimal Separating Hyperplane, which lands for front foot and with metapedes
The Optimal Separating Hyperplane function on ground.As shown in Figure 6A, include the following steps：
Step 601, corresponding multiple data segments during sole of user during running contacts to earth are obtained；
Step 602, the characteristic for carrying out gait analysis is determined from multiple data segments；
Step 603, characteristic is input to Optimal Separating Hyperplane function, obtains output result；
Step 604, gait monitoring result of user during running is determined according to output result.
The present embodiment is similar with embodiment described in above-mentioned Fig. 5, by the way that characteristic is input to Optimal Separating Hyperplane function, obtains
To output as a result, determining gait monitoring result of user during running according to output result, to which gait monitoring be asked
Gait monitoring can also be continuously improved for binary classification problems in topic conversion by using the Optimal Separating Hyperplane that mass data obtains
As a result accuracy.
Fig. 7 A show the flow diagram of the gait monitoring method of an exemplary embodiment seven according to the present invention, Fig. 7 B
Show the model schematic based on decision tree of Fig. 7 A illustrated embodiments；The present embodiment with the mathematical model trained be based on
It illustrates for the model of decision tree, when the model to decision tree carries out parameter training, determines in training data
Including front foot data and metapedes data, wherein training data is used to train the model of decision tree；By in training data
The threshold value of each node in front foot data and metapedes data iterative estimate decision tree, wherein the relative theory of decision tree refers to
The prior art, the application are not described in detail；As shown in Figure 7 A, include the following steps：
Step 701, corresponding multiple data segments during sole of user during running contacts to earth are obtained；
Step 702, the characteristic for carrying out gait analysis is determined from multiple data segments；
Step 703, characteristic threshold value corresponding with the node is compared in each node of decision tree, with true
Determine the next node in decision tree；
Step 704, the final leaf node according to characteristic in decision tree determines step of user during running
State monitoring result.
As shown in Figure 7 B, it is illustrated so that decision tree includes 5 nodes as an example, it, will at first node 711
The corresponding characteristic of first node first threshold corresponding with the node is compared, next in decision tree to determine
Node is second node 712 or third node 713, if the corresponding characteristic of first node is more than the first threshold
Value, then decision enters second node 712, if the corresponding characteristic of first node is less than or equal to first threshold,
Then decision enters third node 713.If into second node 712, by second 712 corresponding characteristic of node
Second threshold corresponding with the node is compared, and is that front foot lands or metapedes lands according to comparison result determination；If into
Enter third node 713, then compares 713 corresponding characteristic of third node third threshold value corresponding with the node
Compared with according to comparison result decision being landed into the 4th node 714 or metapedes, if decision enters the 4th node
714, then the 4th 714 corresponding characteristic of node the 4th threshold value corresponding with the node is compared, is tied according to comparing
Decisive and resolute plan is landed into the 5th node 715 or metapedes, if decision enters the 5th node 715, by the 5th
715 corresponding characteristic of node the 5th threshold value corresponding with the node is compared, and is that front foot according to comparison result decision
Ground or metapedes land.
It will be appreciated by persons skilled in the art that being only by be trained to decision tree one shown in Fig. 7 B
The example of decision tree, wherein the corresponding threshold value of each node, and determine the magnitude relationship of data characteristics and corresponding threshold value
What decision went out in turn redirects result (judging next time that is, the output judging result or present node of present node jump to), all
It is obtained by repetitive exercise, therefore Fig. 7 B are only one and illustratively describe that the limitation to the application can not be formed.
By foregoing description it is found that the present embodiment is similar with embodiment described in above-mentioned Fig. 5 and Fig. 6 A, existed according to characteristic
Final leaf node in decision tree determines gait monitoring result of user during running, to turn gait monitoring problem
Binary classification problems have been changed to, gait monitoring can also be continuously improved by using the decision tree threshold parameter that mass data obtains
As a result accuracy.
By above-described embodiment, the application can make ordinary person without using the large size in the laboratory of sports science research
Analytical instrument carries out the gait analysis of profession, and under the premise of no professional person instructs, can correctly recognize oneself
Gait, and adjustment appropriate is made according to the result of gait analysis, to improve running horizontal and avoid sport injury to reach.
Corresponding to above-mentioned gait monitoring method, the application also proposed shown in Fig. 8 exemplary according to the one of the application
The schematic configuration diagram of the wearable device of embodiment.Referring to FIG. 8, in hardware view, which includes processor, interior
Portion's bus, network interface, memory and nonvolatile memory are also possible that the required hardware of other business certainly.Place
Reason device is formed real from then operation in corresponding computer program to memory is read in nonvolatile memory on logic level
Existing gait monitoring device.Certainly, other than software realization mode, other realization methods, such as logic is not precluded in the application
The mode etc. of device or software and hardware combining, that is to say, that the executive agent of following process flow is not limited to each logic
Unit can also be hardware or logical device.
Fig. 9 is the structural schematic diagram according to the gait monitoring device of one example embodiment of the present invention；As shown in figure 9,
The gait monitoring device may include：First acquisition module 91, the first determining module 92, the first computing module 93.Wherein：
First acquisition module 91, for obtaining corresponding multiple data during sole of user during running contacts to earth
Section；
First determining module 92, multiple data segments for being got from the first acquisition module 91 are determined for carrying out gait
The characteristic of analysis；
First computing module 93, the characteristic for determining the first determining module 92 are input to the mathematical modulo trained
In type, gait monitoring result of user during running is obtained.
Figure 10 shows the structural schematic diagram of gait monitoring device in accordance with an alternative illustrative embodiment of the present invention, device
It may also include：
Second determining module 94 is contacted to earth process for obtaining sole of the user in running during in the first acquisition module 91
In before corresponding multiple data segments, determine whether user is in running state；
Second acquisition module 95 obtains user and exists if determining that user is in running state for the second determining module 94
Sensing data during running；
Preprocessing module 96, the sensing data for being got to the second acquisition module 95 are pre-processed, are used
Corresponding multiple data segments during sole of family during running contacts to earth, so that the first acquisition module 91 obtains.
In one embodiment, preprocessing module can 96 include：
Filter unit 961, for carrying out signal filtering to sensing data；
First detection unit 962, the user detected for 961 filtered sensing data of detection filter unit are running
Corresponding wave crest and trough during each step sole during step contacts to earth；
First determination unit 963, the corresponding wave crest of each step for being detected according to first detection unit 962 and trough
Determine sole of user during running contact to earth during corresponding multiple data segments.
In one embodiment, the first determining module 92 may include：
It is corresponding in frequency domain to obtain multiple data segments for carrying out FFT transform to multiple data segments for FFT transform unit 921
Energy spectrum；
Second determination unit 922, for being determined on the energy spectrum that FFT transform unit 921 obtains for indicating energy point
The characteristic of cloth, wherein the characteristic of Energy distribution is for carrying out gait analysis.
In one embodiment, the first determining module 92 may include：
Third determination unit 923 is respectively right for being determined within each data segment corresponding period to multiple data segments
Mean value, wave crest feature, trough feature, peak valley concussion value and the peak valley spacing answered, wherein mean value, wave crest feature, trough are special
Sign, peak valley concussion value and peak valley spacing are for carrying out gait analysis.
Figure 11 shows the structural schematic diagram of gait monitoring device in accordance with a further exemplary embodiment of the present invention, one
In embodiment, the mathematical model trained is Distance conformability degree model, and device may also include：
Third determining module 97, for when similarity model of adjusting the distance carries out parameter training, determining institute in training data
Including front foot data and metapedes data, wherein training data is for training Distance conformability degree model；
First training module 98, for calculating separately the front foot data of the determination of third determining module 97 in Distance conformability degree mould
The metapedes data that corresponding first distributed constant and third determining module determine in type are corresponding the in Distance conformability degree model
Two distributed constants；
First computing module 93 includes：
First computing unit 931, for characteristic to be carried out phase with the first distributed constant and the second distributed constant respectively
It is calculated like degree, obtains the first similarity and the second similarity；
4th determination unit 932, the first similarity and second for being calculated according to the first computing unit 931 are similar
Degree determines gait monitoring result of user during running.
In one embodiment, the mathematical model trained is supporting vector machine model, and device may also include：
4th determining module 99, for when carrying out parameter training to supporting vector machine model, determining institute in training data
Including front foot data and metapedes data, wherein training data be used for Training Support Vector Machines model；
Second training module 11, front foot data and metapedes data for being determined according to the 4th determining module 99 train one
A maximum Optimal Separating Hyperplane function of spacing distance made between front foot and corresponding two classifications of metapedes, Optimal Separating Hyperplane function
It lands the classification function to land with metapedes as front foot；
First computing module 93 may include：
Second computing unit 933 obtains output result for characteristic to be input to Optimal Separating Hyperplane function；
5th determination unit 934, the output result for being calculated according to the second computing unit 933 determine that user is running
Gait monitoring result during step.
In one embodiment, the mathematical model trained is the model based on decision tree, and device may also include：
5th determining module 12, for when the model to decision tree carries out parameter training, determining in training data and being wrapped
The front foot data and metapedes data contained, wherein training data is used to train the model of decision tree；
Third training module 13, the front foot data and metapedes in training data for being determined by the 5th determining module 12
The threshold value of each node in data iterative estimate decision tree；
First computing module 93 may include：
Comparing unit 935 carries out characteristic threshold value corresponding with the node for each node in decision tree
Compare, to determine the next node in decision tree；
6th determination unit 936, for determining that user is running according to final leaf node of the characteristic in decision tree
Gait monitoring result during step.
By above-described embodiment, ordinary person can be made without using the large-scale analyzer in the laboratory of sports science research
Device carries out the gait analysis of profession, and under the premise of no professional person instructs, can correctly recognize the gait of oneself,
And adjustment appropriate is made according to the result of gait analysis, to reach raising running level and avoid sport injury.
Those skilled in the art after considering the specification and implementing the invention disclosed here, will readily occur to its of the application
Its embodiment.This application is intended to cover any variations, uses, or adaptations of the application, these modifications, purposes or
Person's adaptive change follows the general principle of the application and includes the undocumented common knowledge in the art of the application
Or conventional techniques.The description and examples are only to be considered as illustrative, and the true scope and spirit of the application are by following
Claim is pointed out.
It should also be noted that, the terms "include", "comprise" or its any other variant are intended to nonexcludability
Including so that process, method, commodity or equipment including a series of elements include not only those elements, but also wrap
Include other elements that are not explicitly listed, or further include for this process, method, commodity or equipment intrinsic want
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that wanted including described
There is also other identical elements in the process of element, method, commodity or equipment.
The foregoing is merely the preferred embodiments of the application, not limiting the application, all essences in the application
With within principle, any modification, equivalent substitution, improvement and etc. done should be included within the scope of the application protection god.
Claims (17)
1. a kind of gait monitoring method, which is characterized in that the method includes：
Corresponding data segment during user's sole of each step during running contacts to earth is obtained, obtains the user entire
The step number run during running and the corresponding multiple data segments of the step number；
The characteristic for carrying out gait analysis is determined from the multiple data segment；
The characteristic is input in the mathematical model trained, gait monitoring of user during running is obtained
It heelstrike lands with full sole as a result, the gait includes forward roll, the rear foot.
2. according to the method described in claim 1, it is characterized in that, the sole for obtaining user's each step during running
Corresponding data segment during contacting to earth obtains the step number and step number correspondence that the user is run during entire running
Multiple data segments the step of before, the method further includes：
Determine whether user is in running state；
If user is in running state, sensing data of user during running is obtained；
The sensing data is pre-processed, during obtaining the user sole of each step contacting to earth during running
Corresponding data segment obtains the step number and the corresponding multiple data of the step number that the user is run during entire running
Section.
3. according to the method described in claim 2, it is characterized in that, described pre-process the sensing data, obtain
Corresponding data segment during user sole of each step during running contacts to earth, obtains the user and is entirely running
The step number run in the process and the corresponding multiple data segments of the step number, including：
Signal filtering is carried out to the sensing data；
Each step sole of the user detected in the sensing data after detection filter during running contacts to earth process
In corresponding wave crest and trough；
Determine that sole of user during running is right during contacting to earth according to the corresponding wave crest of each step and trough
The multiple data segments answered.
4. according to the method described in claim 1, it is characterized in that, described determine from the multiple data segment for carrying out gait
The characteristic of analysis, including：
FFT transform is carried out to the multiple data segment, obtains the multiple data segment in the corresponding energy spectrum of frequency domain；
The characteristic for indicating Energy distribution is determined on the energy spectrum, wherein the characteristic of the Energy distribution
For carrying out gait analysis.
5. according to the method described in claim 1, it is characterized in that, described determine from the multiple data segment for carrying out gait
The characteristic of analysis, including：
Corresponding mean value, wave crest feature, wave are determined within each data segment corresponding period to the multiple data segment
Paddy feature, peak valley concussion value and peak valley spacing, wherein the mean value, wave crest feature, trough feature, peak valley concussion value and
Peak valley spacing is for carrying out gait analysis.
6. according to the method described in claim 1, it is characterized in that, the mathematical model trained is Distance conformability degree mould
Type, the method further include：
When carrying out parameter training to the Distance conformability degree model, front foot data and metapedes included in training data are determined
Data, wherein the training data is for training the Distance conformability degree model；
Calculate separately the front foot data corresponding first distributed constant and metapedes number in the Distance conformability degree model
According to corresponding second distributed constant in the Distance conformability degree model；
It is described that the characteristic is input in the mathematical model trained, obtain gait of user during running
Monitoring result, including：
The characteristic is subjected to similarity calculation with first distributed constant and second distributed constant respectively, is obtained
First similarity and the second similarity；
Gait monitoring result of user during running is determined according to first similarity and second similarity.
7. according to the method described in claim 1, it is characterized in that, the mathematical model trained is support vector machines mould
Type, the method further include：
When carrying out parameter training to the supporting vector machine model, front foot data and metapedes included in training data are determined
Data, wherein the training data is for training the supporting vector machine model；
Training one according to the front foot data and the metapedes data makes between front foot and corresponding two classifications of metapedes
The maximum Optimal Separating Hyperplane function of spacing distance, the Optimal Separating Hyperplane function land the classification letter to land with metapedes as front foot
Number；
It is described that the characteristic is input in the mathematical model trained, obtain gait of user during running
Monitoring result, including：
The characteristic is input to the Optimal Separating Hyperplane function, obtains output result；
Gait monitoring result of user during running is determined according to the output result.
8. according to the method described in claim 1, it is characterized in that, the mathematical model trained is the mould based on decision tree
Type, the method further include：
When the model to the decision tree carries out parameter training, front foot data and metapedes number included in training data are determined
According to, wherein the training data is used to train the model of the decision tree；
By each being saved in decision tree described in front foot data in the training data and the metapedes data iterative estimate
The threshold value of point；
It is described that the characteristic is input in the mathematical model trained, obtain gait of user during running
Monitoring result, including：
Characteristic threshold value corresponding with the node is compared in each node of the decision tree, to determine
State the next node in decision tree；
Step of user during running is determined according to final leaf node of the characteristic in the decision tree
State monitoring result.
9. a kind of gait monitoring device, which is characterized in that described device includes：
First acquisition module, for obtaining corresponding data segment during user's sole of each step during running contacts to earth,
Obtain the step number and the corresponding multiple data segments of the step number that the user is run during entire running；
First determining module, the multiple data segment for being got from first acquisition module are determined for carrying out gait
The characteristic of analysis；
First computing module, the characteristic for determining first determining module are input to the mathematical modulo trained
In type, obtain gait monitoring result of user during running, the gait include forward roll, the rear foot heelstrike
It lands with full sole.
10. device according to claim 9, which is characterized in that described device further includes：
Second determining module is right during sole of the user in running during contacts to earth for being obtained in first acquisition module
Before the multiple data segments answered, determine whether user is in running state；
Second acquisition module obtains user and exists if determining that the user is in running state for second determining module
Sensing data during running；
Preprocessing module, the sensing data for being got to second acquisition module pre-process, and obtain institute
State corresponding multiple data segments during sole of user during running contacts to earth.
11. device according to claim 10, which is characterized in that the preprocessing module includes：
Filter unit, for carrying out signal filtering to the sensing data；
First detection unit is being run for detecting the user that the sensing data after the filtering unit filters detects
Corresponding wave crest and trough during each step sole in the process contacts to earth；
First determination unit, the corresponding wave crest of each step and trough for being detected according to the first detection unit are true
Corresponding multiple data segments during fixed sole of user during running contacts to earth.
12. device according to claim 9, which is characterized in that first determining module includes：
FFT transform unit obtains the multiple data segment and is corresponded in frequency domain for carrying out FFT transform to the multiple data segment
Energy spectrum；
Second determination unit, for being determined on the energy spectrum that the FFT transform unit obtains for indicating Energy distribution
Characteristic, wherein the characteristic of the Energy distribution is for carrying out gait analysis.
13. device according to claim 9, which is characterized in that first determining module includes：
Third determination unit, for determination to be corresponding within each data segment corresponding period to the multiple data segment
Mean value, wave crest feature, trough feature, peak valley concussion value and peak valley spacing, wherein the mean value, wave crest feature, trough are special
Sign, peak valley concussion value and peak valley spacing are for carrying out gait analysis.
14. device according to claim 9, which is characterized in that the mathematical model trained is Distance conformability degree mould
Type, described device further include：
Third determining module, for when carrying out parameter training to the Distance conformability degree model, determining in training data and being wrapped
The front foot data and metapedes data contained, wherein the training data is for training the Distance conformability degree model；
First training module, for calculating separately front foot data that the third determining module determines described apart from similar
The metapedes data that corresponding first distributed constant and the third determining module determine in degree model are described apart from similar
Spend corresponding second distributed constant in model；
First computing module includes：
First computing unit, for by the characteristic respectively with first distributed constant and second distributed constant into
Row similarity calculation obtains the first similarity and the second similarity；
4th determination unit, first similarity for being calculated according to first computing unit and second phase
Gait monitoring result of user during running is determined like degree.
15. device according to claim 9, which is characterized in that the mathematical model trained is support vector machines mould
Type, described device further include：
4th determining module, for when carrying out parameter training to the supporting vector machine model, determining in training data and being wrapped
The front foot data and metapedes data contained, wherein the training data is for training the supporting vector machine model；
Second training module, the front foot data and the metapedes data for being determined according to the 4th determining module are trained
Go out a maximum Optimal Separating Hyperplane function of spacing distance made between front foot and corresponding two classifications of metapedes, the classification is super
Planar function lands the classification function to land with metapedes as front foot；
First computing module includes：
Second computing unit obtains output result for the characteristic to be input to the Optimal Separating Hyperplane function；
5th determination unit, the output result for being calculated according to second computing unit determine that the user exists
Gait monitoring result during running.
16. device according to claim 9, which is characterized in that the mathematical model trained is based on decision tree
Model, described device further include：
5th determining module, for when the model to the decision tree carries out parameter training, determining included in training data
Front foot data and metapedes data, wherein the training data is used to train the model of the decision tree；
Third training module, front foot data in the training data for being determined by the 5th determining module with
The threshold value of each node in decision tree described in the metapedes data iterative estimate；
First computing module includes：
Comparing unit carries out characteristic threshold value corresponding with the node for each node in the decision tree
Compare, with the next node in the determination decision tree；
6th determination unit, for determining the user according to final leaf node of the characteristic in the decision tree
Gait monitoring result during running.
17. a kind of wearable device, which is characterized in that the wearable device includes：
Processor；Memory for storing the processor-executable instruction；
Wherein, the processor, is configured as：
Corresponding data segment during user's sole of each step during running contacts to earth is obtained, obtains the user entire
The step number run during running and the corresponding multiple data segments of the step number；
The characteristic for carrying out gait analysis is determined from the multiple data segment；
The characteristic is input in the mathematical model trained, gait monitoring of user during running is obtained
It heelstrike lands with full sole as a result, the gait includes forward roll, the rear foot.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510561183.9A CN105268171B (en) | 2015-09-06 | 2015-09-06 | gait monitoring method, device and wearable device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510561183.9A CN105268171B (en) | 2015-09-06 | 2015-09-06 | gait monitoring method, device and wearable device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105268171A CN105268171A (en) | 2016-01-27 |
| CN105268171B true CN105268171B (en) | 2018-09-18 |
Family
ID=55138540
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510561183.9A Active CN105268171B (en) | 2015-09-06 | 2015-09-06 | gait monitoring method, device and wearable device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105268171B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022066095A1 (en) * | 2020-09-25 | 2022-03-31 | Walkbeat Ab | System and method for analyzing gait in humans |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106237604A (en) * | 2016-08-31 | 2016-12-21 | 歌尔股份有限公司 | Wearable device and the method utilizing its monitoring kinestate |
| CN108404382A (en) * | 2018-02-24 | 2018-08-17 | 上海康斐信息技术有限公司 | A kind of algorithm and system of adaptive training plan generation |
| CN109256187A (en) * | 2018-08-27 | 2019-01-22 | 爱动超越人工智能科技（北京）有限责任公司 | A kind of running posture suggesting system for wearing and method based on multimode motion sensor data |
| CN109260673A (en) * | 2018-11-27 | 2019-01-25 | 北京羽扇智信息科技有限公司 | A kind of movement method of counting, device, equipment and storage medium |
| CN110575177B (en) * | 2019-09-10 | 2020-12-15 | 电子科技大学 | Gait classification and quantification method based on Mahalanobis distance |
| CN116637352A (en) * | 2020-03-31 | 2023-08-25 | 荣耀终端有限公司 | Running gesture detection method and wearable device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1705956A (en) * | 2002-10-29 | 2005-12-07 | 索尼株式会社 | Gait waveform feature extracting method and individual identification system |
| CN101477618A (en) * | 2008-12-18 | 2009-07-08 | 上海交通大学 | Process for pedestrian step gesture periodic automatic extraction from video |
| CN102670207A (en) * | 2012-05-15 | 2012-09-19 | 北京大学 | Gait analysis method based on plantar pressure |
| JP2013022188A (en) * | 2011-07-20 | 2013-02-04 | Nippon Telegr & Teleph Corp <Ntt> | Gait analyzing method, gait analyzer, and program of the same |
| CN103876756A (en) * | 2014-04-18 | 2014-06-25 | 南京工程学院 | Lower limb power-assisted exoskeleton robot gait pattern identification method and system |
| CN104299003A (en) * | 2014-11-02 | 2015-01-21 | 天津理工大学 | Gait recognition method based on similar rule Gaussian kernel function classifier |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2804596B1 (en) * | 2000-02-04 | 2002-10-04 | Agronomique Inst Nat Rech | METHOD FOR THE ANALYSIS OF HUMAN LOCOMOTION IRREGULARITIES |
| TWI498846B (en) * | 2012-07-06 | 2015-09-01 | Univ Nat Cheng Kung | Gait analysis method and gait analysis system |
-
2015
- 2015-09-06 CN CN201510561183.9A patent/CN105268171B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1705956A (en) * | 2002-10-29 | 2005-12-07 | 索尼株式会社 | Gait waveform feature extracting method and individual identification system |
| CN101477618A (en) * | 2008-12-18 | 2009-07-08 | 上海交通大学 | Process for pedestrian step gesture periodic automatic extraction from video |
| JP2013022188A (en) * | 2011-07-20 | 2013-02-04 | Nippon Telegr & Teleph Corp <Ntt> | Gait analyzing method, gait analyzer, and program of the same |
| CN102670207A (en) * | 2012-05-15 | 2012-09-19 | 北京大学 | Gait analysis method based on plantar pressure |
| CN103876756A (en) * | 2014-04-18 | 2014-06-25 | 南京工程学院 | Lower limb power-assisted exoskeleton robot gait pattern identification method and system |
| CN104299003A (en) * | 2014-11-02 | 2015-01-21 | 天津理工大学 | Gait recognition method based on similar rule Gaussian kernel function classifier |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022066095A1 (en) * | 2020-09-25 | 2022-03-31 | Walkbeat Ab | System and method for analyzing gait in humans |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105268171A (en) | 2016-01-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105268171B (en) | gait monitoring method, device and wearable device | |
| CN105411593B (en) | Method and apparatus for identifying gait task | |
| Awais et al. | Physical activity classification for elderly people in free-living conditions | |
| US10503967B2 (en) | Fast behavior and abnormality detection | |
| Uddin et al. | Random forests based recognition of human activities and postural transitions on smartphone | |
| US20140032165A1 (en) | Apparatus, system, and method for automatic identification of sensor placement | |
| CN108021888A (en) | A kind of fall detection method | |
| Ferrero et al. | On gait recognition with smartphone accelerometer | |
| Kawsar et al. | A novel activity detection system using plantar pressure sensors and smartphone | |
| JP6943287B2 (en) | Biometric information processing equipment, biometric information processing systems, biometric information processing methods, and programs | |
| Zhang | Application of human motion recognition utilizing deep learning and smart wearable device in sports | |
| Barri Khojasteh et al. | Evaluation of a wrist-based wearable fall detection method | |
| Aiolli et al. | ClimbTheWorld: Real-time stairstep counting to increase physical activity | |
| Youn et al. | Wearable sensor-based biometric gait classification algorithm using WEKA | |
| Huang et al. | Recent trends in wearable device used to detect freezing of gait and falls in people with Parkinson’s disease: A systematic review | |
| CN109800733A (en) | Data processing method and device, electronic equipment | |
| Adel et al. | Gait-based Person Identification using Multiple Inertial Sensors. | |
| Dohnálek et al. | Human activity recognition on raw sensor data via sparse approximation | |
| Esseiva et al. | Feet fidgeting detection based on accelerometers using decision tree learning and gradient boosting | |
| KR20210022375A (en) | Apparatus and method for identifying individuals by performing discriminant analysis for various detection information | |
| KR20210046121A (en) | Apparatus and method for identify patients with parkinson's disease and patients with podarthritis by performing neural network analysis by various detection information | |
| Donahue et al. | Estimation of ground reaction force waveforms during fixed pace running outside the laboratory | |
| Zhang et al. | How to record the amount of exercise automatically? A general real-time recognition and counting approach for repetitive activities | |
| Wang et al. | Human motion phase segmentation based on three new features | |
| KR20140020440A (en) | Activity diagnosis apparatus and method of the same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |