CN108932504A - Identity identifying method, device, electronic equipment and storage medium - Google Patents
Identity identifying method, device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN108932504A CN108932504A CN201810822151.3A CN201810822151A CN108932504A CN 108932504 A CN108932504 A CN 108932504A CN 201810822151 A CN201810822151 A CN 201810822151A CN 108932504 A CN108932504 A CN 108932504A
- Authority
- CN
- China
- Prior art keywords
- gait
- feature
- pattern
- acceleration signal
- step state
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
- G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The present embodiments relate to identity identifying technology field, a kind of identity identifying method, device, electronic equipment and storage medium are provided, which comprises obtain the step state acceleration signal for the acceleration transducer acquisition being set to user;Gait cycle detection is carried out to step state acceleration signal, obtains multiple gait cycles;Multiple gait cycles are divided into multiple gait segmentation sections, and schema extraction is carried out to each gait segmentation section, obtain the corresponding gait pattern of each gait segmentation section;Characteristic vector pickup is carried out to each gait pattern, obtains the gait feature of each gait pattern;According to the gait feature of each gait pattern, identification model is constructed, to confirm user identity.Compared with prior art, the embodiment of the present invention obtains the step state acceleration signal of different user and determines gait cycle, can satisfy the individual difference of different user, realizes that the user identity based on Gait Recognition precisely identifies.
Description
Technical field
The present embodiments relate to identity identifying technology field, in particular to a kind of identity identifying method, device,
Electronic equipment and storage medium.
Background technique
Gait refers to the posture of walking of people, everyone gait is unique and uniqueness, therefore gait is a kind of packet
The behavior biological characteristic of identity information containing people.In recent years, with the development of micro-electronic mechanical system technique, low cost micromation
Acceleration transducer rapidly develop and be widely used in various mobile wearable devices, this makes the acquisition of body gait information
And the authentication based on gait is widely studied.
Currently, the research method of Gait Recognition can be divided into three kinds: based on machine vision, based on pressure sensor and being based on
Wearable sensors.Wherein, based on the gait pattern that machine vision is during being walked using video camera capture sequence of user
Then image realizes authentication using image matching algorithm.However, vulnerable to light, block, distance etc. influence;Based on pressure
Sensor is easily affected by using the gait feature of pressure sensor capture user;Based on wearable sensors
The gait signal of people is extracted by being worn on the acceleration transducer of human body different location and realize user identity authentication or
Identification.Compared with based on machine vision method and pressure sensor method, the method based on wearable sensors can more directly
Verily reflect the gait characteristic of people, but do not account for the leg speed difference of user when detecting in the prior art, ignores
The individual difference of user, therefore there is a problem of that detection is not accurate.
Summary of the invention
The embodiment of the present invention is designed to provide a kind of identity identifying method, device, electronic equipment and storage medium, uses
To realize the accurate identification of user identity based on Gait Recognition.
To achieve the goals above, technical solution used in the embodiment of the present invention is as follows:
In a first aspect, the embodiment of the invention provides a kind of identity identifying methods, which comprises acquisition is set to use
The step state acceleration signal of acceleration transducer acquisition with family;Gait cycle inspection is carried out to the step state acceleration signal
It surveys, obtains multiple gait cycles;Multiple gait cycles are divided into multiple gait segmentation sections, and Duan Junjin is divided to each gait
Row schema extraction obtains the corresponding gait pattern of each gait segmentation section;Characteristic vector pickup is carried out to each gait pattern,
Obtain the gait feature of each gait pattern;According to the gait feature of each gait pattern, identification model is constructed, with confirmation
User identity.
Second aspect, the embodiment of the invention also provides a kind of identification authentication system, described device includes obtaining module, inspection
Survey module, schema extraction module, characteristic vector pickup module and model construction module.Wherein, module is obtained for obtaining setting
The step state acceleration signal of acceleration transducer acquisition with user;Detection module is used for the step state acceleration signal
Gait cycle detection is carried out, multiple gait cycles are obtained;Schema extraction module is used to multiple gait cycles being divided into multiple steps
State divides section, and carries out schema extraction to each gait segmentation section, obtains the corresponding gait pattern of each gait segmentation section;It is special
For sign vector extraction module for carrying out characteristic vector pickup to each gait pattern, the gait for obtaining each gait pattern is special
Sign;Model construction module constructs identification model, for the gait feature according to each gait pattern to confirm user's body
Part.
The third aspect, the embodiment of the invention also provides a kind of electronic equipment, the electronic equipment and user's is wearable
Communication of mobile terminal connects, and is provided with acceleration transducer, the electronic equipment further include: one in the wearable mobile terminal
A or multiple processors;Memory, for storing one or more programs, when one or more of programs by one or
When multiple processors execute, so that one or more of processors realize above-mentioned identity identifying method.
Fourth aspect, the embodiment of the invention also provides a kind of computer readable storage mediums, are stored thereon with computer
Program, the computer program realize above-mentioned identity identifying method when being executed by processor.
Compared with the prior art, a kind of identity identifying method provided in an embodiment of the present invention, device, electronic equipment and storage are situated between
Matter firstly, obtaining the step state acceleration signal for the acceleration transducer acquisition being set to user, and is believed step state acceleration
Number carry out gait cycle detect to obtain multiple gait cycles;Then, multiple gait cycles are divided into multiple gaits and divide section, and
Schema extraction is carried out to each gait segmentation section, obtains the corresponding gait pattern of each gait segmentation section;MFCC is recycled to calculate
Method carries out characteristic vector pickup to each gait pattern, obtains the gait feature of each gait pattern;Finally according to each step
The gait feature of morphotype formula constructs identification model, to confirm user identity.Compared with prior art, the embodiment of the present invention
It obtains the step state acceleration signal of different user and determines gait cycle, can satisfy the individual difference of different user, realize
User identity based on Gait Recognition precisely identifies.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, preferred embodiment is cited below particularly, and cooperate
Appended attached drawing, is described in detail below.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 shows the box that electronic equipment provided in an embodiment of the present invention and acceleration transducer interact and illustrates
Figure.
Fig. 2 shows the block diagrams of electronic equipment provided in an embodiment of the present invention.
Fig. 3 shows identity identifying method flow chart provided in an embodiment of the present invention.
Fig. 4 shows the step state acceleration signal schematic representation of acceleration transducer acquisition.
Fig. 5 is Z axis acceleration information schematic diagram in the step state acceleration signal shown in Fig. 4.
Fig. 6 is the autocorrelation signal schematic diagram of the step state acceleration signal shown in Fig. 4.
Fig. 7 is that the gait of the step state acceleration signal shown in Fig. 4 divides section schematic diagram,
Fig. 8 shows gait cycle test experience result schematic diagram.
Fig. 9 shows gait segmentation schematic diagram data.
Figure 10 shows gait pattern and extracts experimental result schematic diagram.
Figure 11 shows the block diagram of identification authentication system provided in an embodiment of the present invention.
Icon: 10- electronic equipment;20- acceleration transducer;101- memory;102- storage control;103- processing
Device;104- internal interface;105- communication unit;200- identification authentication system;201- obtains module;202- preprocessing module;
203- detection module;204- schema extraction module;205- characteristic vector pickup module;206- model construction module.
Specific embodiment
Below in conjunction with attached drawing in the embodiment of the present invention, technical solution in the embodiment of the present invention carries out clear, complete
Ground description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be arranged and be designed with a variety of different configurations herein.Cause
This, is not intended to limit claimed invention to the detailed description of the embodiment of the present invention provided in the accompanying drawings below
Range, but it is merely representative of selected embodiment of the invention.Based on the embodiment of the present invention, those skilled in the art are not doing
Every other embodiment obtained under the premise of creative work out, shall fall within the protection scope of the present invention.
It should also be noted that similar label and letter indicate similar terms in following attached drawing, therefore, once a certain Xiang Yi
It is defined in a attached drawing, does not then need that it is further defined and explained in subsequent attached drawing.Meanwhile of the invention
In description, term " first ", " second " etc. are only used for distinguishing description, are not understood to indicate or imply relative importance.
Wearable device is fast-developing in recent years, is widely used to network communication, tele-medicine, the outer rehabilitation of hospital, shifting
A variety of application fields such as dynamic payment, interactive game, the function of becoming stronger day by day is (for example, information communication, mobile payment, personal silver
Row, positioning etc.) and storage a large amount of personal information (for example, communication information, physiological health parameter, picture, voice, video etc.) make
Obtaining safety and secret protection becomes an important need of wearable device.Currently, being directed to the safety and privacy of wearable device
Protection mainly includes cipher encrypting method and biological authentication method, wherein there are passwords to be easily stolen for cipher encrypting method, is easily broken
The problems such as solution, cannot guarantee the safety of wearable device well, in addition, increasing with wearable device, password is easy quilt
User forgets, makes troubles for the use of wearable device.Biological authentication method includes fingerprint recognition and recognition of face, secrecy
Property it is preferable, but need special fingerprint or human face scanning component, it is complicated for operation, at high cost and not can be carried out continuous certification.Thus
As it can be seen that existing still have defect for the safety of wearable device and the method for secret protection.
Identity identifying method based on Gait Recognition is mainly to add to the gait of the accelerometer acquisition in wearable device
Speed signal is analyzed to carry out user identity authentication, is cooperated on one's own initiative with user is not needed and is identified more accurate spy
Property.Since gait is a kind of behavior biological characteristic of identity information comprising people, therefore the identity identifying method based on Gait Recognition
It is more accurate.
Fig. 1 is please referred to, Fig. 1 shows electronic equipment 10 provided in an embodiment of the present invention and at least one acceleration transducer
20 block diagrams interacted.Electronic equipment 10 is communicated with acceleration transducer 20, and electronic equipment 10 is accelerated by obtaining
The step state acceleration signal that sensor 20 acquires is spent, and Gait Recognition is carried out to realize that user identity is recognized to step state acceleration signal
Card.Electronic equipment 10 may be, but not limited to, the intelligent electronic devices such as desktop computer, laptop, the behaviour of electronic equipment 10
It may be, but not limited to, IOS (iPhone operating system) system, Windows system etc. as system.Acceleration passes
Sensor 20 can be worn on user by being integrated in wearable device, and wearable device can be mobile phone, bracelet, foot chain
With wearable device being worn on according to oneself actual conditions Deng, user, for example, wrist, arm, chest, waist,
Thigh etc..
Referring to figure 2., Fig. 2 shows the block diagrams of electronic equipment 10 provided in an embodiment of the present invention.Electronic equipment
10 include memory 101, storage control 102, processor 103, internal interface 104 and communication unit 105.The memory
101, storage control 102, processor 103, internal interface 104 and each element of communication unit 105 are direct or indirect between each other
Ground is electrically connected, to realize the transmission or interaction of data.For example, these elements between each other can be total by one or more communication
Line or signal wire, which are realized, to be electrically connected.Identification authentication system 200 includes that at least one can be with software or firmware (firmware)
Form is stored in the memory 101 or is solidificated in the software function module in the operating system of the electronic equipment 10.Institute
It states processor 103 to be used to execute the executable module stored in memory 101, such as the identification authentication system 200 includes
Software function module or computer program.
Wherein, memory 101 may be, but not limited to, random access memory (Random Access Memory,
RAM), read-only memory (Read Only Memory, ROM), programmable read only memory (Programmable Read-Only
Memory, PROM), erasable read-only memory (Erasable Programmable Read-Only Memory, EPROM),
Electricallyerasable ROM (EEROM) (Electric Erasable Programmable Read-Only Memory, EEPROM) etc..
Processor 103 can be a kind of IC chip, have signal handling capacity.Above-mentioned processor 103 can be with
It is general processor, including central processing unit (Central Processing Unit, CPU), network processing unit (Network
Processor, NP), speech processor and video processor etc.;Can also be digital signal processor, specific integrated circuit,
Field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components.
It may be implemented or execute disclosed each method, step and the logic diagram in the embodiment of the present invention.General processor can be
Microprocessor or the processor 103 are also possible to any conventional processor etc..
The internal interface 104 is used to couple processor 103 and memory 101 for various input/output devices.?
In some embodiments, internal interface 104, processor 103 and storage control 102 can be realized in one single chip.At it
In his some examples, they can be realized by independent chip respectively.
Communication unit 105 is used for through the company of foundation between wireless network and the acceleration transducer 20 being set to user
It connects, to realize that electronic equipment 10 passes through wireless network sending and receiving data.
First embodiment
Referring to figure 3., Fig. 3 shows identity identifying method flow chart provided in an embodiment of the present invention.Identity identifying method
The following steps are included:
Step S101 obtains the step state acceleration signal for the acceleration transducer acquisition being set to user.
In embodiments of the present invention, acceleration transducer 20 can be worn on user's body by being integrated in wearable device
On, acceleration transducer 20 can synchronization acquire three quadrature axis X-axis, Y-axis, Z axis acceleration information, that is,
It says, in the process of walking, acceleration transducer 20 exports X-axis acceleration information, Y-axis acceleration information and Z axis acceleration to user
Data.Electronic equipment 10 obtains the step state acceleration signal that acceleration transducer 20 acquires, electronic equipment by communication unit 105
10 can obtain the step state acceleration signal that acceleration transducer 20 acquires by communication unit 105, and step state acceleration signal can
To include multiple sample points (for example, 1200), each sample point includes X-axis acceleration information, Y-axis acceleration information and Z
Axle acceleration data.
Step S102 pre-processes step state acceleration signal, to reduce the noise in step state acceleration signal.
In embodiments of the present invention, since 20 sampling clock of acceleration transducer integrated in wearable device is unstable,
Cause the time interval between 2 adjacent continuous sample points inconsistent, therefore can be using the method for linear interpolation to having obtained
The step state acceleration signal taken is handled, so that it is guaranteed that the time interval between sample point is fixed.
In addition, acceleration transducer 20 is during acquiring step state acceleration signal, due to sensor position is not fixed,
The influence of ground inequality factor, the step state acceleration signal that electronic equipment 10 is got inevitably contain much noise,
It therefore can be using the noise point weakened in step state acceleration signal based on the noise remove method that multilevel wavelet decomposes and reconstructs
Step state acceleration signal comprising noise specifically can be carried out 2 grades of small wavelength-divisions using db6 wavelet orthogonal basis function by amount
Solution, the whole wavelet coefficients retained under large scale low resolution can set the wavelet coefficient under each scale high-resolution
One threshold value, amplitude all set 0 lower than the wavelet coefficient of the threshold value, and the wavelet coefficient higher than the threshold value completely retains or received
Contracting processing, then the wavelet coefficient obtained after processing is reconstructed using wavelet inverse transformation and recovers effective step state acceleration letter
Number, it reduces by acquisition environment bring signal noise.
Step S103 carries out gait cycle detection to step state acceleration signal, obtains multiple gait cycles.
In embodiments of the present invention, a gait cycle can be heel first contacts ground to same heel again
The time interval between ground is contacted, posture, speed, step-length walked due to each user etc. is different, therefore each user
Gait cycle is uniquely that different users corresponds to different gait cycles, and acceleration transducer 20 collects under normal conditions
Step state acceleration signal include multiple gait cycles, it is therefore desirable to pretreated step state acceleration signal carry out gait week
Phase detection, Fig. 4 are shown the step state acceleration signal schematic representation of the acquisition of acceleration transducer 20, can be accelerated from Fig. 4 with gait
The Z axis acceleration information spent in signal has stronger periodicity, therefore can use Z axis acceleration information to carry out gait cycle
Detection, specific method may include:
Firstly, the multiple minimum points for obtaining Z axis acceleration information in step state acceleration signal specifically arbitrarily take
The acceleration value of the acceleration value of the sample point sample point adjacent with its left and right is compared by one sample point, if the sample
The acceleration value of the acceleration value of the point sample point more adjacent than with its left and right is all small, it is determined that the sample point is minimum point, is pressed
According to same method, all sample points in Z axis acceleration information are traversed, multiple minimums in Z axis acceleration information are obtained
Point, for example, Z axis acceleration information schematic diagram of the Fig. 5 for preceding 600 sample points in the step state acceleration signal shown in Fig. 4, Fig. 5
The middle sample point with " * " label is minimum point.
Secondly, the acceleration value according to each minimum point, filters out the noise spot in multiple minimum points, due to user's row
The influence of the factors such as weight during walking may have some noise spots in the multiple minimum points got, need this
A little noise spots filter out.Specifically, acceleration value that can first according to multiple minimum points, calculates the school of these minimum points
Quasi- difference and mean value;Again according to the calibration difference of these minimum points and mean value, a threshold value is determined to filter out multiple minimums
Noise spot in point, threshold value can be calculated according to formula Threshold=mean+0.5*std, wherein Threshold table
Show threshold value, std indicates standard deviation, and mean indicates mean value;The noise spot in multiple minimum points is finally found out according to the threshold value, it will
The minimum point that acceleration value is greater than the threshold value is determined as noise spot, and filters out all noise spots found out.
Next, calculating the auto-correlation coefficient of Z axis acceleration information, and determine to estimate step-length according to auto-correlation coefficient,
Specifically, the auto-correlation coefficient of each sample point in Z axis acceleration information can be first calculated, and to obtained auto-correlation coefficient
It is normalized to obtain autocorrelation signal;Then the disposal of gentle filter is carried out to autocorrelation signal, filters out autocorrelation signal
In noise, calculate the auto-correlation coefficient of the Z axis acceleration information shown in Fig. 5, obtained autocorrelation signal is as shown in Figure 6;Most
Afterwards, the interval in autocorrelation signal obtained in the previous step between first minimum point and third minimum point is determined as walking
The state period estimates step-length L.
Finally, multiple gait cycles are extracted according to filtering out multiple minimum points after noise spot and estimating step-length, due to
The minimum point that second step obtains is not necessarily all the starting point or terminating point of gait cycle, therefore needs to obtain second step minimum
Value point makees further screening, and then minimum point and estimates the starting point and ending point that step-length L finds out each gait cycle.Below
It is illustrated for detecting first gait cycle:
First, calculate the first interval d1 between first minimum point and second minimum point, that is, statistics the
Sample points between one minimum point and second minimum point, and using the first interval d1 and estimate the relationship between step-length
To look for the starting point and ending point of first gait cycle.Specifically, ifI.e. first interval d1, which is less than, estimates step
LongThen determine that first minimum point and second minimum point are not the starting point of gait cycle, compares the at this time
The acceleration value of one minimum point and second minimum point, and reject the biggish minimum point of acceleration value;IfI.e. first interval d1, which is greater than, estimates step-lengthAnd it is less than or equal to estimate step-lengthThen determine first
Minimum point and second minimum point are also not the terminating point of gait cycle;
Second, on the basis of first minimum point, calculate between third minimum point and second minimum point
Second interval d2, and using the second interval d2 and estimate the relationship between step-length and look for the starting point and end of first gait cycle
Stop.Specifically, ifIt is biggish then to reject acceleration value in third minimum point and second minimum point
Minimum point;IfSecond minimum point is then rejected, then detects to obtain first gait cycle by above method
Starting point and ending point be respectively first minimum point and third minimum point;
Third, ifI.e. first interval d1, which is greater than, estimates step-lengthBetween between adjacent minimum point
Every not exceeding a gait cycle, therefore step-length is estimated when the first interval d1 is greater thanWhen, the starting of first gait cycle
Point and terminating point are respectively first minimum point and second minimum point.
It is rising for next gait cycle with the terminating point of previous gait cycle after detecting first gait cycle
Initial point loops through each minimum point according to above method, the terminating point of next gait cycle is searched, until detecting
Some gait cycles.
Multiple gait cycles are divided into multiple gaits and divide section, and carried out to each gait segmentation section by step S104
Schema extraction obtains the corresponding gait pattern of each gait segmentation section.
In embodiments of the present invention, to step state acceleration signal carry out gait cycle detection, obtain multiple gait cycles it
It afterwards, is that a gait divides section with 4 gait cycles, the overlapping of setting 50% between adjacent 2 gaits segmentation section will be multiple
Gait cycle is divided into multiple gait segmentation sections, and segmentation result is as shown in Figure 7.
In embodiments of the present invention, after obtaining multiple gait segmentation sections, mode is carried out to each gait segmentation section and is mentioned
It takes, obtains the corresponding gait pattern of each gait segmentation section, specifically, firstly, to the corresponding gait of each gait segmentation section
Acceleration signal carries out data fusion, obtains corresponding first fused data of each gait segmentation section and the second fused data,
In, the first fused data is the fusion of X-axis acceleration information and Y-axis acceleration information, and the first fused data is that X-axis accelerates degree
According to, the fusion of Y-axis acceleration information and Z axis acceleration information.Utilize X-axis, Y-axis and the Z axis in each gait segmentation section, benefit
Use formulaAndResettle two axis MXYAxis and MXYZAxis, the first fused data
For MXYAxle acceleration data, the second fused data are MXYZAxle acceleration data;Then, by each gait segmentation section corresponding the
One fused data, the second fused data and Z axis acceleration information are combined, and obtain the corresponding gait mould of each gait segmentation section
Formula, that is to say, that choose Z axis, the M in each gait segmentation sectionXYAxis and MXYZThe data of axis are corresponding as each gait segmentation section
Gait pattern, as P=[aZ,aMXY,aMXYZ]。
Step S105 carries out characteristic vector pickup to each gait pattern, obtains the gait feature of each gait pattern.
In embodiments of the present invention, each gait pattern includes that the first fused data, the second fused data and Z axis accelerate
Degree evidence, i.e. P=[aZ,aMXY,aMXYZ].The specific method that characteristic vector pickup is carried out to each gait pattern may is that
Firstly, using MFCC algorithm to the first fused data, the second fused data and the Z axis acceleration information in each gait pattern point
It is not calculated, obtains corresponding fisrt feature feature vector, second feature vector and third feature vector, that is to say, that benefit
Z axis acceleration information, M in each gait pattern P are extracted with MFCC algorithmXYAxle acceleration data and MXYZIn axle acceleration data
MFCC coefficient, fisrt feature feature vector be Z axis acceleration information MFCC coefficient, second feature vector be MXYAxle acceleration
The MFCC coefficient of data, third feature vector are MXYZThe MFCC coefficient of axle acceleration data, fisrt feature feature vector, second
Feature vector and third feature vector are 18 dimension MFCC vectors;Then, by the fisrt feature feature vector, second feature to
Amount and third feature vector are merged, and obtain the gait feature of each gait pattern, that is to say, that by Z axis, MXYAxis and MXYZ
The corresponding 3 18 dimensions MFCC Vector Fusion of axis is got up, and the feature vector of one 54 dimension, the gait feature of each gait pattern are constituted
It is all the feature vector of one 54 dimension.
Step S106 constructs identification model, according to the gait feature of each gait pattern to confirm user identity.
In embodiments of the present invention, to prevent calculation amount excessive, gait feature obtained above is carried out at PCA dimensionality reduction
Reason, meanwhile, in order to promote the convergence precision of identification model, the gait feature after dimension-reduction treatment is normalized,
Then using sorting algorithm (for example, support vector machines, neural network, K value arest neighbors etc.) common in machine learning to gait
Feature is classified, to confirm user identity.Specifically, firstly, gait feature according to each gait pattern, constructs instruction
Practice data set and test data set, since each step state acceleration signal that acceleration transducer 20 acquires has phase in storage
The record label answered, it is that record label is used to characterize for which secondary data recorded, therefore can will record number is all of even number
Record number is all gait features of odd number as test data set as training dataset by gait feature;Then, by institute
State training dataset input Machine learning classifiers, train identification model, Machine learning classifiers can be support to
Amount machine classifier, neural network classifier, k nearest neighbor classification device etc.;Finally, test data set is inputted trained identification
Model outputs test data and concentrates the classification of each gait feature.When carrying out user identity authentication, user is carried out in advance
Gait sample training stores the gait feature extracted, when carrying out user identity authentication, using machine learning algorithm
Identification certification is carried out to gait, classifier output is+1 if it exists, then passes through authentication;If the output of all classifiers is-
1, then do not pass through authentication.
The identity identifying method that embodiment provides in order to better illustrate the present invention has compared with the higher identification of the prior art
Precision is illustrated below by way of an experiment, specific as follows:
Step state acceleration number is acquired using acceleration transducer 20 built-in in Google's Android HTC Nexus One mobile phone
According to, the data of totally 38 subjects, wherein 28 people of male, 10 people of women, average age 23 years old to 28 years old.In data acquisition
Mobile phone is put into the trousers pocket of subject with constant direction and is set as SENSOR_DELAY_ on Android SDK
FASTEST mode, sample rate are about 27Hz.Every volunteer acquires 15 times or more data, every time when record data all
Write down be which time record data, and tagged stored.
Gait cycle test experience is carried out first, and experimental results are shown in figure 8, with markd point is just gait in Fig. 8
The starting point or terminating point in period, it can be seen that all starting point, terminating point are and adjacent minimum all on minimum point
The interval of value point is uniform, and also in the contiguous range for estimating step-length, while having very strong periodicity;
Then gait pattern is extracted, since the acceleration information of X-axis and Y-axis is unstable, is divided in section by each gait
X-axis, Y-axis and Z axis establish two axis MXYAxis and MXYZAxis, gait are divided just there are five the data of axis in section, and Fig. 9 is this five
The acceleration information of axis.Z axis, M are used in embodiments of the present inventionXYAxis and MXYZThe acceleration information of axis as a gait pattern,
Figure 10 is the acceleration information under a gait pattern.
Gait feature identification is finally carried out, chooses the data of 30 subjects in data set, wherein 22 people of male, women 8
People shares 559 record data, and wherein even-times record data have 287 times, and odd-times records data 272 times, the gait of storage
Feature has 5026, and wherein even-times record data have 2596, and odd-times record data have 2430, records number with even-times
According to for training dataset, it is test data set that odd-times, which records data, and training dataset is imported in each different classifier
In (for example, support vector machines, neural network, K value arest neighbors), identification model is obtained, then test data set is imported into
Forecasting recognition is carried out in the identification model, so as to obtain accuracy of identification, classification results table is as shown in table 1.
1 classification results table of table
Sorting algorithm | Accuracy rate |
Support vector machines | 90.33% |
Neural network | 86.42% |
K value is closest | 77.65% |
It can see by table 1, the accuracy rate highest of support vector machines.
Compared with prior art, the embodiment of the present invention has the advantage that
Firstly, seeking auto-correlation coefficient to step state acceleration signal and finding out minimum point, according to the first of auto-correlation coefficient
A and third minimum point estimates step-length, and rejects the minimum point of non-gait cycle according to step-length is estimated, to find standard
The starting point and ending point of true gait cycle;
Secondly as the acceleration information of X-axis and Y-axis is unstable, X-axis, Y-axis and the Z in section are divided by each gait
Axis establishes two axis MXYAxis and MXYZAxis eliminates this influence, and extracts Z axis, MXYAxis and MXYZThe acceleration information conduct of axis
Gait pattern, to obtain metastable gait pattern data.
Finally, calculating Z axis, M using MFCC algorithmXYAxis and MXYZThe MFCC coefficient of the acceleration information of axis is as gait
Feature can more comprehensively be truly reflected gait feature, and than traditional, only extraction gait feature is more smart in time domain or frequency domain
Really.
Second embodiment
Figure 11 is please referred to, Figure 11 shows the block diagram of identification authentication system 200 provided in an embodiment of the present invention.Body
Part authentication device 200 include obtain module 201, preprocessing module 202, detection module 203, schema extraction module 204, feature to
Measure extraction module 205 and model construction module 206.
Module 201 is obtained, for obtaining the step state acceleration signal of the acceleration transducer being set to user acquisition.
Preprocessing module 202, for being pre-processed to step state acceleration signal, to reduce in step state acceleration signal
Noise.
Detection module 203 obtains multiple gait cycles for carrying out gait cycle detection to step state acceleration signal.
In embodiments of the present invention, step state acceleration signal includes Z axis acceleration information, and detection module 203 is specifically used for
Obtain multiple minimum points in Z axis acceleration information;According to the acceleration value of each minimum point, multiple minimum points are filtered out
In noise spot;The auto-correlation coefficient of Z axis acceleration information is counted, and determines to estimate step-length according to auto-correlation coefficient;According to filter
Except multiple minimum points after noise spot and step-length is estimated, extracts multiple gait cycles.
Schema extraction module 204 divides section for multiple gait cycles to be divided into multiple gaits, and to each gait point
It cuts section and carries out schema extraction, obtain the corresponding gait pattern of each gait segmentation section.
In embodiments of the present invention, schema extraction module 204, specifically for adding to the corresponding gait of each gait segmentation section
Speed signal carries out data fusion, obtains corresponding first fused data of each gait segmentation section and the second fused data;It will be every
A gait is divided corresponding first fused data of section, the second fused data and Z axis acceleration information and is combined, and each step is obtained
State divides the corresponding gait pattern of section.
Characteristic vector pickup module 205 obtains each gait for carrying out characteristic vector pickup to each gait pattern
The gait feature of mode.
In embodiments of the present invention, each gait pattern includes that the first fused data, the second fused data and Z axis accelerate
Degree evidence, characteristic vector pickup module 205, specifically for merging number to first in each gait pattern using MFCC algorithm
It is respectively calculated according to, the second fused data and Z axis acceleration information, obtains corresponding fisrt feature feature vector, second special
Levy vector and third feature vector;The fisrt feature feature vector, second feature vector and third feature vector are melted
It closes, obtains the gait feature of each gait pattern.
Model construction module 206 constructs identification model, for the gait feature according to each gait pattern with true
Recognize user identity.
In embodiments of the present invention, schema extraction module 204, specifically for the gait feature according to each gait pattern,
Construct training dataset and test data set;The training dataset is inputted into Machine learning classifiers, trains the body
Part identification model;The test data set is inputted into the trained identification model, outputing test data, it is each to concentrate
The classification of gait feature, to confirm user identity.
The embodiment of the present invention further discloses a kind of computer readable storage medium, is stored thereon with computer program, described
The identity identifying method that present invention discloses is realized when computer program is executed by processor 103.
In conclusion a kind of identity identifying method, device, electronic equipment and storage medium provided in an embodiment of the present invention,
The described method includes: obtaining the step state acceleration signal for the acceleration transducer acquisition being set to user;Gait is accelerated
It spends signal and carries out gait cycle detection, obtain multiple gait cycles;Multiple gait cycles are divided into multiple gait segmentation sections, and
Schema extraction is carried out to each gait segmentation section, obtains the corresponding gait pattern of each gait segmentation section;To each gait mould
Formula carries out characteristic vector pickup, obtains the gait feature of each gait pattern;According to the gait feature of each gait pattern, structure
Identification model is built, to confirm user identity.Compared with prior art, the gait that the embodiment of the present invention obtains different user adds
Speed signal simultaneously determines gait cycle, can satisfy the individual difference of different user, realizes user's body based on Gait Recognition
Part precisely identification.
In several embodiments provided herein, it should be understood that disclosed device and method can also pass through
Other modes are realized.The apparatus embodiments described above are merely exemplary, for example, flow chart and block diagram in attached drawing
Show the device of multiple embodiments according to the present invention, the architectural framework in the cards of method and computer program product,
Function and operation.In this regard, each box in flowchart or block diagram can represent the one of a module, section or code
Part, a part of the module, section or code, which includes that one or more is for implementing the specified logical function, to be held
Row instruction.It should also be noted that function marked in the box can also be to be different from some implementations as replacement
The sequence marked in attached drawing occurs.For example, two continuous boxes can actually be basically executed in parallel, they are sometimes
It can execute in the opposite order, this depends on the function involved.It is also noted that every in block diagram and or flow chart
The combination of box in a box and block diagram and or flow chart can use the dedicated base for executing defined function or movement
It realizes, or can realize using a combination of dedicated hardware and computer instructions in the system of hardware.
In addition, each functional module in each embodiment of the present invention can integrate one independent portion of formation together
Point, it is also possible to modules individualism, an independent part can also be integrated to form with two or more modules.
It, can be with if the function is realized and when sold or used as an independent product in the form of software function module
It is stored in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially in other words
The part of the part that contributes to existing technology or the technical solution can be embodied in the form of software products, the meter
Calculation machine software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be a
People's computer, server or network equipment etc.) it performs all or part of the steps of the method described in the various embodiments of the present invention.
And storage medium above-mentioned includes: that USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), arbitrary access are deposited
The various media that can store program code such as reservoir (RAM, Random Access Memory), magnetic or disk.It needs
Illustrate, herein, relational terms such as first and second and the like be used merely to by an entity or operation with
Another entity or operation distinguish, and without necessarily requiring or implying between these entities or operation, there are any this realities
The relationship or sequence on border.Moreover, the terms "include", "comprise" or its any other variant are intended to the packet of nonexcludability
Contain, so that the process, method, article or equipment for including a series of elements not only includes those elements, but also including
Other elements that are not explicitly listed, or further include for elements inherent to such a process, method, article, or device.
In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including the element
Process, method, article or equipment in there is also other identical elements.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.It should also be noted that similar label and letter exist
Similar terms are indicated in following attached drawing, therefore, once being defined in a certain Xiang Yi attached drawing, are then not required in subsequent attached drawing
It is further defined and explained.
Claims (10)
1. a kind of identity identifying method, which is characterized in that the described method includes:
Obtain the step state acceleration signal for the acceleration transducer acquisition being set to user;
Gait cycle detection is carried out to the step state acceleration signal, obtains multiple gait cycles;
Multiple gait cycles are divided into multiple gait segmentation sections, and schema extraction is carried out to each gait segmentation section, are obtained
Each gait divides the corresponding gait pattern of section;
Characteristic vector pickup is carried out to each gait pattern, obtains the gait feature of each gait pattern;
According to the gait feature of each gait pattern, identification model is constructed, to confirm user identity.
2. the method as described in claim 1, which is characterized in that the step state acceleration signal includes Z axis acceleration information, institute
The step of stating and gait cycle detection carried out to the step state acceleration signal, obtaining multiple gait cycles, comprising:
Obtain multiple minimum points in the Z axis acceleration information;
According to the acceleration value of each minimum point, the noise spot in the multiple minimum point is filtered out;
The auto-correlation coefficient of the Z axis acceleration information is calculated, and determines to estimate step-length according to the auto-correlation coefficient;
According to filter out multiple minimum points after noise spot and it is described estimate step-length, extract multiple gait cycles.
3. method according to claim 2, which is characterized in that it is described that schema extraction is carried out to each gait segmentation section, it obtains
The step of gait pattern corresponding to each gait segmentation section, comprising:
Data fusion is carried out to the corresponding step state acceleration signal of each gait segmentation section, it is corresponding to obtain each gait segmentation section
First fused data and the second fused data;
Corresponding first fused data of each gait segmentation section, the second fused data and Z axis acceleration information are combined, obtained
To the corresponding gait pattern of each gait segmentation section.
4. method as claimed in claim 3, which is characterized in that each gait pattern includes the first fused data, second melts
Data and Z axis acceleration information are closed, it is described that feature extraction is carried out to each gait pattern, obtain the gait of each gait pattern
The step of feature, comprising:
Using MFCC algorithm to the first fused data, the second fused data and the Z axis acceleration information in each gait pattern point
It is not calculated, obtains corresponding fisrt feature feature vector, second feature vector and third feature vector;
The fisrt feature feature vector, second feature vector and third feature vector are merged, each gait mould is obtained
The gait feature of formula.
5. the method as described in claim 1, which is characterized in that the gait feature according to each gait pattern constructs body
Part identification model, the step of to confirm user identity, comprising:
According to the gait feature of each gait pattern, training dataset and test data set are constructed;
The training dataset is inputted into Machine learning classifiers, trains the identification model;
The test data set is inputted into the trained identification model, outputs test data and concentrates each gait feature
Classification, to confirm user identity.
6. the method as described in claim 1, which is characterized in that the acceleration transducer being set to user that obtains is adopted
After the step of step state acceleration signal of collection, the method also includes:
The step state acceleration signal is pre-processed, to reduce the noise in the step state acceleration signal.
7. a kind of identification authentication system, which is characterized in that described device includes:
Module is obtained, for obtaining the step state acceleration signal of the acceleration transducer being set to user acquisition;
Detection module obtains multiple gait cycles for carrying out gait cycle detection to the step state acceleration signal;
Schema extraction module divides section for multiple gait cycles to be divided into multiple gaits, and divides Duan Jun to each gait
Schema extraction is carried out, the corresponding gait pattern of each gait segmentation section is obtained;
Characteristic vector pickup module obtains each gait pattern for carrying out characteristic vector pickup to each gait pattern
Gait feature;
Model construction module constructs identification model, for the gait feature according to each gait pattern to confirm user's body
Part.
8. device as claimed in claim 7, which is characterized in that described device further include:
Preprocessing module, for being pre-processed to the step state acceleration signal, to reduce in the step state acceleration signal
Noise.
9. a kind of electronic equipment, which is characterized in that the electronic equipment and the wearable communication of mobile terminal of user connect, described
Acceleration transducer, the electronic equipment are provided in wearable mobile terminal further include:
One or more processors;
Memory, for storing one or more programs, when one or more of programs are by one or more of processors
When execution, so that one or more of processors realize such as method of any of claims 1-6.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program quilt
Such as method of any of claims 1-6 is realized when processor executes.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810822151.3A CN108932504A (en) | 2018-07-24 | 2018-07-24 | Identity identifying method, device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810822151.3A CN108932504A (en) | 2018-07-24 | 2018-07-24 | Identity identifying method, device, electronic equipment and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108932504A true CN108932504A (en) | 2018-12-04 |
Family
ID=64444345
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810822151.3A Pending CN108932504A (en) | 2018-07-24 | 2018-07-24 | Identity identifying method, device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108932504A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111544005A (en) * | 2020-05-15 | 2020-08-18 | 中国科学院自动化研究所 | Parkinson's disease dyskinesia quantification and identification method based on support vector machine |
WO2020211459A1 (en) * | 2019-04-15 | 2020-10-22 | 歌尔股份有限公司 | Use object determining method, device, and storage medium |
CN113065468A (en) * | 2021-04-02 | 2021-07-02 | 西安电子科技大学 | Gait authentication method based on user coordinate system and GRU network |
CN113140051A (en) * | 2020-01-20 | 2021-07-20 | 上海依图信息技术有限公司 | Attendance checking method and device, electronic equipment and storage medium |
CN113591552A (en) * | 2021-06-18 | 2021-11-02 | 新绎健康科技有限公司 | Method and system for identity recognition based on gait acceleration data |
CN115906032A (en) * | 2023-02-20 | 2023-04-04 | 之江实验室 | Recognition model correction method and device and storage medium |
CN117421605A (en) * | 2023-10-27 | 2024-01-19 | 绍兴清研微科技有限公司 | Gait recognition method and system based on block chain technology |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101695445A (en) * | 2009-10-29 | 2010-04-21 | 浙江大学 | Acceleration transducer-based gait identification method |
US20130179112A1 (en) * | 2012-01-09 | 2013-07-11 | Honeywell International Inc. | Robust method for signal segmentation for motion classification in personal navigation |
CN106056013A (en) * | 2016-06-03 | 2016-10-26 | 深圳市金立通信设备有限公司 | Terminal burglary prevention method and terminal |
CN106850955A (en) * | 2016-12-20 | 2017-06-13 | 陕西尚品信息科技有限公司 | A kind of mobile phone identity verification method based on Gait Recognition |
CN107016346A (en) * | 2017-03-09 | 2017-08-04 | 中国科学院计算技术研究所 | gait identification method and system |
CN107122704A (en) * | 2017-03-16 | 2017-09-01 | 华南理工大学 | A kind of gait recognition method based on motion sensor |
CN107831907A (en) * | 2017-12-07 | 2018-03-23 | 深圳先进技术研究院 | Identity identifying method and device based on Gait Recognition |
CN108022248A (en) * | 2016-11-03 | 2018-05-11 | 北京航空航天大学 | A kind of lower limb gait rehabilitation assessment system of view-based access control model collecting device |
-
2018
- 2018-07-24 CN CN201810822151.3A patent/CN108932504A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101695445A (en) * | 2009-10-29 | 2010-04-21 | 浙江大学 | Acceleration transducer-based gait identification method |
US20130179112A1 (en) * | 2012-01-09 | 2013-07-11 | Honeywell International Inc. | Robust method for signal segmentation for motion classification in personal navigation |
CN106056013A (en) * | 2016-06-03 | 2016-10-26 | 深圳市金立通信设备有限公司 | Terminal burglary prevention method and terminal |
CN108022248A (en) * | 2016-11-03 | 2018-05-11 | 北京航空航天大学 | A kind of lower limb gait rehabilitation assessment system of view-based access control model collecting device |
CN106850955A (en) * | 2016-12-20 | 2017-06-13 | 陕西尚品信息科技有限公司 | A kind of mobile phone identity verification method based on Gait Recognition |
CN107016346A (en) * | 2017-03-09 | 2017-08-04 | 中国科学院计算技术研究所 | gait identification method and system |
CN107122704A (en) * | 2017-03-16 | 2017-09-01 | 华南理工大学 | A kind of gait recognition method based on motion sensor |
CN107831907A (en) * | 2017-12-07 | 2018-03-23 | 深圳先进技术研究院 | Identity identifying method and device based on Gait Recognition |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020211459A1 (en) * | 2019-04-15 | 2020-10-22 | 歌尔股份有限公司 | Use object determining method, device, and storage medium |
CN113140051A (en) * | 2020-01-20 | 2021-07-20 | 上海依图信息技术有限公司 | Attendance checking method and device, electronic equipment and storage medium |
CN111544005A (en) * | 2020-05-15 | 2020-08-18 | 中国科学院自动化研究所 | Parkinson's disease dyskinesia quantification and identification method based on support vector machine |
CN111544005B (en) * | 2020-05-15 | 2022-03-08 | 中国科学院自动化研究所 | Parkinson's disease dyskinesia quantification and identification method based on support vector machine |
CN113065468A (en) * | 2021-04-02 | 2021-07-02 | 西安电子科技大学 | Gait authentication method based on user coordinate system and GRU network |
CN113065468B (en) * | 2021-04-02 | 2023-04-07 | 西安电子科技大学 | Gait authentication method based on user coordinate system and GRU network |
CN113591552A (en) * | 2021-06-18 | 2021-11-02 | 新绎健康科技有限公司 | Method and system for identity recognition based on gait acceleration data |
CN115906032A (en) * | 2023-02-20 | 2023-04-04 | 之江实验室 | Recognition model correction method and device and storage medium |
CN117421605A (en) * | 2023-10-27 | 2024-01-19 | 绍兴清研微科技有限公司 | Gait recognition method and system based on block chain technology |
CN117421605B (en) * | 2023-10-27 | 2024-04-30 | 绍兴清研微科技有限公司 | Gait recognition method and system based on block chain technology |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108932504A (en) | Identity identifying method, device, electronic equipment and storage medium | |
US10130307B2 (en) | Electrocardiogram (ECG) authentication method and apparatus | |
CN107506684B (en) | Gait recognition method and device | |
Thang et al. | Gait identification using accelerometer on mobile phone | |
Muaaz et al. | An analysis of different approaches to gait recognition using cell phone based accelerometers | |
Hoang et al. | Adaptive cross-device gait recognition using a mobile accelerometer | |
US20160294825A1 (en) | Supplementing biometric identification with device identification | |
CN107831907A (en) | Identity identifying method and device based on Gait Recognition | |
Sarkar et al. | Deep feature-based face detection on mobile devices | |
Kovač et al. | Human skeleton model based dynamic features for walking speed invariant gait recognition | |
Qin et al. | A fuzzy authentication system based on neural network learning and extreme value statistics | |
KR20050025927A (en) | The pupil detection method and shape descriptor extraction method for a iris recognition, iris feature extraction apparatus and method, and iris recognition system and method using its | |
CN111178155A (en) | Gait feature extraction and gait recognition method based on inertial sensor | |
Benzaoui et al. | Experiments and improvements of ear recognition based on local texture descriptors | |
Al-Naffakh et al. | Unobtrusive gait recognition using smartwatches | |
Ferrero et al. | On gait recognition with smartphone accelerometer | |
CN102629320A (en) | Ordinal measurement statistical description face recognition method based on feature level | |
CN110826610A (en) | Method and system for intelligently detecting whether dressed clothes of personnel are standard | |
CN107403161A (en) | Biological feather recognition method and device | |
Muaaz et al. | Accelerometer based gait recognition using adapted gaussian mixture models | |
Derawi | Smartphones and biometrics: Gait and activity recognition | |
Cherifi et al. | Robust multimodal biometric authentication on IoT device through ear shape and arm gesture | |
Abate et al. | Smartphone enabled person authentication based on ear biometrics and arm gesture | |
Muaaz et al. | Cross pocket gait authentication using mobile phone based accelerometer sensor | |
Derawi et al. | Fusion of gait and ECG for biometric user authentication |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181204 |