CN109492703A - A kind of recognition methods of gait, system and terminal device - Google Patents
A kind of recognition methods of gait, system and terminal device Download PDFInfo
- Publication number
- CN109492703A CN109492703A CN201811407341.5A CN201811407341A CN109492703A CN 109492703 A CN109492703 A CN 109492703A CN 201811407341 A CN201811407341 A CN 201811407341A CN 109492703 A CN109492703 A CN 109492703A
- Authority
- CN
- China
- Prior art keywords
- gait
- walking
- data
- hidden markov
- walking data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/112—Gait analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
- G06F18/295—Markov models or related models, e.g. semi-Markov models; Markov random fields; Networks embedding Markov models
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Pathology (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Heart & Thoracic Surgery (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Dentistry (AREA)
- Physiology (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The present invention is suitable for computer application technology, provide a kind of recognition methods of gait, system and terminal device, the recognition methods includes: to obtain walking sample data and walking data to be identified, according to preset model parameter, standard hidden Markov model is initialized, based on walking sample data, standard hidden Markov model after initialization is trained, based on viterbi algorithm, utilize the standard hidden Markov model and walking data after training, determine gait split path, gait split path includes several gait spliting nodes, according to gait spliting node, walking data are divided into several walking data acquisition systems, and determine the corresponding gait of each walking data acquisition system, it is analyzed to obtain gait without using video data, the problem of to be not directed to leakage individual subscriber privacy, it will not It is influenced by dynamic background, improves the accuracy of Gait Recognition.
Description
Technical field
The invention belongs to computer application technologies more particularly to a kind of recognition methods of gait, system and terminal to set
It is standby.
Background technique
With the continuous enhancing of the rapid development of technology of Internet of things and health of people, awareness of safety, by advanced Internet of Things
Technology is applied to human body information and acquires and then provide service for fields such as human health, public safeties, has become promotion intelligence
The key of development of information.In recent years, as one of Human bodys' response key technology, gait Activity recognition technology attracts
The extensive concern of academia.
In the prior art, when carrying out gait Activity recognition, the normal row of user is acquired generally by wearable camera
The video walked obtains the corresponding gait of user, includes the video of user by camera acquisition, be easy sudden and violent by analyzing video
Reveal the privacy of user, and be easy to be influenced by dynamic background, the accuracy of Gait Recognition is reduced, for example, collected portion
Dividing video includes the walking behavior of other pedestrians, therefore, is existed in such a way that the video of analysis camera acquisition determines gait
It is easy leakage privacy of user and the lower problem of Gait Recognition accuracy.
Summary of the invention
In view of this, the embodiment of the invention provides a kind of recognition methods of gait, system and terminal device, it is existing to solve
With the presence of easy leakage privacy of user and gait are known in such a way that the video of analysis camera acquisition determines gait in technology
The lower problem of other accuracy.
The first aspect of the embodiment of the present invention provides a kind of recognition methods of gait, comprising:
Obtain walking sample data and walking data to be identified；
According to preset model parameter, standard hidden Markov model is initialized；
Based on the walking sample data, the standard hidden Markov model after initialization is trained；
Based on viterbi algorithm, using after training standard hidden Markov model and the walking data, determine gait
Split path, the gait split path include several gait spliting nodes；
According to the gait spliting node, the walking data are divided into several walking data acquisition systems, and determine every
The corresponding gait of a walking data acquisition system.
The second aspect of the embodiment of the present invention provides a kind of identifying system of gait, comprising:
Walking data acquisition module, for obtaining walking sample data and walking data to be identified；
Model initialization module, for being initialized to standard hidden Markov model according to preset model parameter；
Model training module, for being based on the walking sample data, to the standard hidden Markov model after initialization
It is trained；
Split path generation module, for be based on viterbi algorithm, using after training standard hidden Markov model and
The walking data determine that gait split path, the gait split path include several gait spliting nodes；
Gait divides module, for according to the gait spliting node, the walking data to be divided into several walkings
Data acquisition system, and determine the corresponding gait of each walking data acquisition system.
The third aspect of the embodiment of the present invention provides a kind of terminal device, including memory, processor and is stored in
In the memory and the computer program that can run on the processor, when the processor executes the computer program
The step of realizing the recognition methods of gait as described above.
The fourth aspect of the embodiment of the present invention provides a kind of computer readable storage medium, the computer-readable storage
Media storage has computer program, and the computer program realizes the recognition methods of gait as described above when being executed by processor
Step.
Existing beneficial effect is the embodiment of the present invention compared with prior art: the embodiment of the present invention obtains walking sample number
The model parameter of standard hidden Markov model is carried out initial according to walking data to be identified according to preset model parameter
Change, the standard hidden Markov model initialized is trained using walking sample data, so that it is determined that hidden out
The model parameter of Markov model is joined using viterbi algorithm according to the corresponding model of standard hidden Markov after training
Several and walking data to be identified, generate gait split path, and the gait spliting node for including according to gait split path is treated
The walking data of identification are split, and obtain several walking data acquisition systems and the corresponding gait of each walking data acquisition system,
It only needs to carry out Gait Recognition using the walking sample data of acquisition and walking data to be identified, be carried out without using video data
Analysis obtains gait, so that the problem of being not directed to leakage individual subscriber privacy, will not be influenced by dynamic background, improves
The accuracy of Gait Recognition.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skill in the art without any creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is the implementation process schematic diagram of the recognition methods of gait provided by one embodiment of the present invention；
Fig. 2 is the specific implementation flow schematic diagram of step S103 in Fig. 1 provided by one embodiment of the present invention；
Fig. 3 is the structural schematic diagram of the identifying system of gait provided by one embodiment of the present invention；
Fig. 4 is the concrete structure schematic diagram of model training module provided by one embodiment of the present invention；
Fig. 5 is the schematic diagram of terminal device provided by one embodiment of the present invention.
Specific embodiment
In being described below, for illustration and not for limitation, the tool of such as particular system structure, technology etc is proposed
Body details, to understand thoroughly the embodiment of the present invention.However, it will be clear to one skilled in the art that there is no these specific
The present invention also may be implemented in the other embodiments of details.In other situations, it omits to well-known system, device, electricity
The detailed description of road and method, in case unnecessary details interferes description of the invention.
In order to illustrate technical solutions according to the invention, the following is a description of specific embodiments.
Embodiment 1:
Fig. 1 shows the implementation process of the recognition methods of the gait of one embodiment of the present of invention offer, and process is described in detail
It is as follows:
In step s101, walking sample data and walking data to be identified are obtained.
In one embodiment of the invention, walking sample data includes at least one walking sample sequence.
In the present embodiment, walking sample sequence includes that sample accelerates degree series and sample angular speed sequence.
Wherein, sample accelerates degree series to contain several sample acceleration informations.
Wherein, sample angular speed sequence contains several sample angular velocity datas.
Wherein, sample angular velocity data is user's angular velocity data that gait is marked.
Wherein, sample acceleration information is user's acceleration information that gait is marked.
In one embodiment of the invention, before step S101, comprising:
1) the frequency adjustment instruction comprising default frequency acquisition is sent to default Inertial Measurement Unit, and frequency adjustment instruction is used
User's walking data are acquired according to default frequency acquisition in the default Inertial Measurement Unit of instruction.
2) user's walking data that default Inertial Measurement Unit is sent are received, walking data to be identified are obtained.
In the present embodiment, it presets Inertial Measurement Unit (Inertial Measurement Unit, IMU) and acquires user
Walking data, walking data include angular velocity data and acceleration information, so that collected data be made to be not directed to user
Privacy concern.
In the present embodiment, higher to the reduction degree of data since the sample frequency of sensor is higher, at the same time to end
The requirement of end equipment is also higher, and terminal device data volume to be treated can also increase with it, and therefore, presets inertia measurement list
The frequency of member acquisition walking data needs the configuration in view of terminal device.
In the present embodiment, the hardware parameter of terminal device is obtained, and from the mapping table prestored, searches the hardware parameter
Corresponding sample frequency obtains the default frequency acquisition.
In the present embodiment, default frequency acquisition is substituting in preset instructions format, generates corresponding frequency adjustment and refers to
It enables, and is sent to default Inertial Measurement Unit, which presets frequency acquisition acquisition user's walking according to this
Data.
In one embodiment of the invention, after step slol, comprising:
1) sliding window size is obtained.
2) it is based on the sliding window size, moving average filter processing is carried out to the walking data.
In the present embodiment, moving average filter algorithm is exactly that the N number of sampled value continuously obtained is regarded as a queue, team
The length of column is fixed as N, and sampling obtains a new data and is put into tail of the queue every time, and loses the data of original head of the queue, in queue
N number of data carry out average calculating operation, so that it may obtain new filter result, wherein N be sliding window size, N is positive integer.
In the present embodiment, since people is in walking, the walking data got have periodically, by walking one by one
State period composition.
In the present embodiment, since the factors such as sensor itself and artificial or nature can generate data noise to walking number
It is periodic therefore, it is necessary to be had the characteristics that according to walking data according to being interfered, good inhibit to make using there is PERIODIC INTERFERENCE
With smoothness is high, and the moving average filter algorithm for being suitable for higher-order of oscillation data is filtered walking data, removes
The noise data of walking data improves the accuracy of walking data.
In step s 102, according to preset model parameter, standard hidden Markov model is initialized.
In the present embodiment, hidden Markov model is general by state-transition matrix A, observation probability matrix B and original state
Rate is distributed π and determines, hidden Markov model λ can be indicated with ternary symbol, i.e. λ=(π, A, B), wherein A, B, π are known as hidden horse
The three elements of Er Kefu model.
In the present embodiment, standard hidden Markov model (Hidden Markov Model, HMM) is without initial
Change and untrained model.
In the present embodiment, preset model parameter includes preset state transfer matrix, default observation probability matrix and presets
The model parameter of standard hidden Markov model is set preset model parameter by initial state probabilities, Ma Erke hidden to standard
Husband's model initializes, for example, setting preset state transfer square for the state-transition matrix of standard hidden Markov model
Battle array.
In step s 103, it is based on walking sample data, the standard hidden Markov model after initialization is trained.
In the present embodiment, walking sample data is to be labelled with the sample angular speed sequence and sample acceleration sequence of gait
Column, can be trained the standard hidden Markov model after initialization by unsupervised-learning algorithm.
In step S104, it is based on viterbi algorithm, using the standard hidden Markov model and walking data after training,
Determine that gait split path, gait split path include several gait spliting nodes.
In the present embodiment, viterbi algorithm really uses dynamic Programming hidden Markov model forecasting problem, that is, uses
Maximum probability path (optimal path) is asked in Dynamic Programming.
In the present embodiment, walking data include angular speed sequence and acceleration degree series, the hidden Ma Erke of standard after training
The model parameter (that is, state-transition matrix A, observation probability matrix B and initial state probabilities are distributed π) of husband's model it has been determined that
Walking data are input in the hidden Markov model after training, several local maximums is exported, is then based on Viterbi
Algorithm determines gait spliting node according to local maximum, i.e., by utilizing the model parameter and angular speed sequence after training
The corresponding probability of each angular speed in angular speed sequence is calculated, local convergence is then carried out, obtains several local poles
Big value, viterbi algorithm determines the position of gait spliting node, i.e., the position of the corresponding data point of each local maximum is true
It is set to gait spliting node, to obtain gait split path, gait split path is to carry out angular speed sequence according to gait
The optimal path of segmentation.
In one embodiment, gait types number is obtained, i.e., a gait cycle includes how many kinds of gait, for example, one
A gait cycle is made of 4 kinds of gaits, can determine a gait cycle may include how many according to gait types number
A walking data acquisition system, i.e. a gait cycle can restrain to obtain how many a local maximums.
By taking a concrete application scene as an example, on the basis of model parameter and angular speed sequence have determined, first calculate
Then the corresponding probability of each angular speed in angular velocity sequence restrains whole probability, restrains several offices out
The quantity of portion's maximum, the local maximum that a gait cycle includes is gait types number, and a gait cycle is by 4 kinds
Gait composition, a kind of spliting node of the corresponding angular speed of each local maximum between gait and another gait, and
A kind of gait and another gait have confirmed specific gait title, for example, a local maximum is pair
The angular speed answered is the 5th angular speed, and the 5th angular speed is the spliting node of gait A Yu gait B.
It can be calculated with the corresponding optimal path of above-mentioned calculating angular speed sequence and accelerate the corresponding optimal path of degree series,
So as to obtain several and accelerate the corresponding walking data acquisition system of degree series and each step to accelerating degree series to be split
Line data set closes corresponding gait.
In step s105, according to gait spliting node, walking data are divided into several walking data acquisition systems, and really
Determine the corresponding gait of each walking data acquisition system.
In one embodiment of the invention, gait includes that heel strike contacts to earth to full vola, full vola contacts to earth to heel
It is liftoff, heel is liftoff arrives heel strike to toe is liftoff and toe is liftoff.
In the present embodiment, a gait cycle be contacted to earth by heel strike to full vola, full vola contact to earth to heel from
Ground, heel are liftoff to be formed to toe is liftoff and toe is liftoff to heel strike, and the circulation of multiple gait cycles constitutes the step of people
Every trade is.
In the present embodiment, each gait spliting node indicates the cut-point of a kind of gait and another gait, and walks
State has determined that, for example, a gait spliting node is that heel strike contacts to earth to full vola and contacts to earth with full vola to foot
Spliting node heeloff.
In the present embodiment, according to gait spliting node, walking data are divided into several walking data acquisition systems, each
Whole corresponding gaits of walking data that walking data acquisition system includes are identical, and since gait spliting node has shown which are
Therefore the cut-point of kind gait and which kind of gait can directly determine out the corresponding gait of each walking data acquisition system, for example,
Walking data include 30 data points altogether, and the 5th data point contacts to earth to full vola for heel strike and contact to earth with full vola to foot
Spliting node heeloff, it is liftoff with the liftoff segmentation liftoff to toe of heel to heel that the 12nd data point is that full vola contacts to earth
Node, then the 1st to 5 data point is walking data acquisition system 1, and the corresponding gait of walking data acquisition system 1 is heel strike to full foot
The palm contacts to earth, and the 6th to 12 data point is walking data acquisition system 2, and the corresponding gait of walking data acquisition system 2 is heel strike to full foot
The palm contacts to earth.
Wherein, spliting node may be embodied in previous set, also may be embodied in the latter set, here, not limiting
System, for example, the 5th data point may be embodied in walking data acquisition system 1, also may be embodied in walking data acquisition system 2.
In the present embodiment, whether the identity or detection user behavior that user can be identified by Gait Recognition advise
Whether model meets the specification gait behavior prestored by the gait behavioral value user behavior of identification user, for example, by sentencing
The disconnected corresponding position of gait spliting node position whether corresponding with the gait spliting node prestored is consistent, so that it is determined that user out
Whether behavior meets specification, and the gait spliting node prestored is the corresponding gait spliting node of gait behavior of specification.
In the present embodiment, walking sample data and walking data to be identified are obtained, according to preset model parameter, to mark
The model parameter of quasi- hidden Markov model is initialized, hidden to the standard initialized using walking sample data
Markov model is trained, so that it is determined that the model parameter of hidden Markov model out, using viterbi algorithm, according to instruction
The corresponding model parameter of standard hidden Markov and walking data to be identified after white silk generate gait split path, according to step
The gait spliting node that state split path includes is split walking data to be identified, obtains several walking data sets
Conjunction and the corresponding gait of each walking data acquisition system, it is only necessary to utilize the walking sample data and walking data to be identified of acquisition
Gait Recognition is carried out, is analyzed to obtain gait without using video data, to be not directed to leakage individual subscriber privacy
Problem will not be influenced by dynamic background, improve the accuracy of Gait Recognition.
The specific implementation flow of the step S103 in Fig. 1 provided Fig. 2 shows one embodiment of the present of invention, process
Details are as follows:
In step s 201, it is based on forward-backward algorithm algorithm, calculates the corresponding sequence probability of each walking sample sequence.
In the present embodiment, in the case where given hidden Markov model initiation parameter model λ=(π, A, B), benefit
The corresponding sequence probability of walking sample sequence is calculated with Forward-backward algorithm, Forward-backward algorithm is by before recurrence calculation
Sequence probability can be efficiently calculated to-backward probability.
In step S202, judge whether the corresponding sequence probability of each walking sample sequence restrains.
In the present embodiment, the corresponding sequence probability of walking sample sequence this calculated comes out with last computation
The corresponding sequence probability of walking sample sequence be compared, that is, calculate both difference, if difference is in preset difference value range
It is interior, then the corresponding sequence probability convergence of the walking sample sequence is judged, if judging within the scope of difference no longer preset difference value
The corresponding sequence probability of the sample sequence is not restrained.
In step S203, sequence probability is not restrained if it exists, then EM algorithm is based on, to the standard after initialization
The corresponding model parameter of hidden Markov model carries out revaluation iteration.
In the present embodiment, sequence probability is not restrained if it exists, then needs to estimate model λ=(π, A, B) parameter, so that
Under the model, the sequence probability of walking sample sequence is maximum, i.e. the sequence probability convergence of walking sample sequence.
In the present embodiment, be based on EM algorithm (EM algorithm), by the corresponding sequence probability of walking sample sequence and
"current" model parameter is substituting in preset model parameter calculation formula, obtains new model parameter, and based on new model parameter
Again be based on forward-backward algorithm algorithm, calculate walking sample sequence sequence probability, if the sequence probability is not restrained still, continue by
New model parameter and new sequence probability substitute into preset model parameter calculation formula, carry out revaluation iteration, until determining
It can make the convergent model parameter of sequence probability.
Wherein, EM algorithm can be Baum welch algorithm.
In step S204, if whole sequence probabilities is restrained, stop to the standard hidden Markov after initialization
Model training.
In the present embodiment, if whole sequence probabilities is restrained, it is determined that training is completed, and is stopped to the mark after initialization
Quasi- hidden Markov model training.
It should be understood that the size of the serial number of each step is not meant that the order of the execution order in above-described embodiment, each process
Execution sequence should be determined by its function and internal logic, the implementation process without coping with the embodiment of the present invention constitutes any limit
It is fixed.
Embodiment 2:
Fig. 3 shows the identifying system 100 of the gait of one embodiment of the present of invention offer, for executing corresponding to Fig. 1
Embodiment in method and step, the identifying system 100 of gait includes:
Walking data acquisition module 110, for obtaining walking sample data and walking data to be identified.
Model initialization module 120, for being carried out to standard hidden Markov model initial according to preset model parameter
Change.
Model training module 130, for be based on walking sample data, to the standard hidden Markov model after initialization into
Row training.
Split path generation module 140 utilizes the standard hidden Markov model after training for being based on viterbi algorithm
With walking data, determine that gait split path, gait split path include several gait spliting nodes.
Gait divides module 150, for according to gait spliting node, walking data to be divided into several walking data sets
It closes, and determines the corresponding gait of each walking data acquisition system.
In one embodiment of the invention, walking data acquisition module 110, comprising:
Window size acquiring unit, for obtaining sliding window size.
Filter processing unit carries out moving average filter processing to walking data for being based on sliding window size.
In one embodiment of the invention, walking data acquisition module 110, further includes:
Frequency instruction transmission unit, for sending the frequency adjustment instruction comprising default frequency acquisition to default inertia measurement
Unit, frequency adjustment instruction are used to indicate default Inertial Measurement Unit and acquire user's walking data according to default frequency acquisition.
Walking data receipt unit, the user's walking data sent for receiving default Inertial Measurement Unit, obtains wait know
Other walking data.
In one embodiment of the invention, walking sample data includes at least one walking sample sequence.
In one embodiment of the invention, gait includes that heel strike contacts to earth to full vola, full vola contacts to earth to heel
It is liftoff, heel is liftoff arrives heel strike to toe is liftoff and toe is liftoff.
As shown in figure 4, in one embodiment, model training module 130 specifically includes:
Sequence probability computing unit 131 calculates the corresponding sequence of each walking sample sequence for being based on forward-backward algorithm algorithm
Column probability.
Judging unit 132 is restrained, for judging whether the corresponding sequence probability of each walking sample sequence restrains.
First processing units 133 are not restrained for sequence probability if it exists, then EM algorithm are based on, to initialization
The corresponding model parameter of standard hidden Markov model afterwards carries out revaluation iteration.
The second processing unit 134 stops if restraining for whole sequence probabilities to the hidden horse of standard after initialization
Er Kefu model training.
In one embodiment, the identifying system 100 of gait further includes other function module/unit, for realizing implementation
Method and step in example 1 in each embodiment.
Embodiment 3:
Fig. 5 is the schematic diagram for the terminal device that one embodiment of the invention provides.As shown in figure 5, the terminal of the embodiment is set
Standby 5 include: processor 50, memory 51 and are stored in the meter that can be run in the memory 51 and on the processor 50
Calculation machine program 52.The processor 50 realizes each embodiment as described in example 1 above when executing the computer program 52
Step, such as step S101 shown in FIG. 1 to step S105.Alternatively, when the processor 50 executes the computer program 52
Realize the function of each module/unit in each system embodiment as described in example 2 above, for example, module 110 shown in Fig. 3 to
150 function.
Illustratively, the computer program 52 can be divided into one or more module/units, it is one or
Multiple module/units are stored in the memory 51, and are executed by the processor 50, to complete the present invention.Described one
A or multiple module/units can be the series of computation machine program instruction section that can complete specific function, which is used for
Implementation procedure of the computer program 52 in the terminal device 5 is described.For example, the computer program 52 can be divided
It is cut into walking data acquisition module, model initialization module, model training module, split path generation module and gait segmentation mould
Block.Each module concrete function is as follows:
Walking data acquisition module, for obtaining walking sample data and walking data to be identified；
Model initialization module, for being initialized to standard hidden Markov model according to preset model parameter；
Model training module, for being based on the walking sample data, to the standard hidden Markov model after initialization
It is trained；
Split path generation module, for be based on viterbi algorithm, using after training standard hidden Markov model and
The walking data determine that gait split path, the gait split path include several gait spliting nodes；
Gait divides module, for according to the gait spliting node, the walking data to be divided into several walkings
Data acquisition system, and determine the corresponding gait of each walking data acquisition system.
The terminal device 5 can be the calculating such as desktop PC, notebook, palm PC and cloud server and set
It is standby.The terminal device 5 may include, but be not limited only to, processor 50, memory 51.It will be understood by those skilled in the art that figure
5 be only the example of terminal device 5, does not constitute the restriction to terminal device 5, may include than illustrating more or fewer portions
Part perhaps combines certain components or different components, such as the terminal device can also include input-output equipment, net
Network access device, bus etc..
Alleged processor 50 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), field programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng.
The memory 51 can be the internal storage unit of the terminal device 5, such as the hard disk or interior of terminal device 5
It deposits.The memory 51 is also possible to the External memory equipment of the terminal device 5, such as be equipped on the terminal device 5
Plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD) card dodge
Deposit card (Flash Card) etc..Further, the memory 51 can also both include the storage inside list of the terminal device 5
Member also includes External memory equipment.The memory 51 is for storing needed for the computer program and the terminal device
Other programs and data.The memory 51 can be also used for temporarily storing the data that has exported or will export.
Embodiment 4:
The embodiment of the invention also provides a kind of computer readable storage medium, computer-readable recording medium storage has meter
Calculation machine program is realized the step in each embodiment as described in example 1 above, such as is schemed when computer program is executed by processor
Step S101 shown in 1 to step S105.Alternatively, realizing when the computer program is executed by processor such as institute in embodiment 2
The function of each module/unit in each system embodiment stated, such as the function of module 110 to 150 shown in Fig. 3.
The computer program can be stored in a computer readable storage medium, and the computer program is by processor
When execution, it can be achieved that the step of above-mentioned each embodiment of the method.Wherein, the computer program includes computer program code,
The computer program code can be source code form, object identification code form, executable file or certain intermediate forms etc..Institute
State computer-readable medium may include: can carry the computer program code any entity or device, recording medium,
USB flash disk, mobile hard disk, magnetic disk, CD, computer storage, read-only memory (ROM, Read-Only Memory), arbitrary access
Memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium etc..It needs
It is bright, the content that the computer-readable medium includes can according in jurisdiction make laws and patent practice requirement into
Row increase and decrease appropriate, such as in certain jurisdictions, according to legislation and patent practice, computer-readable medium do not include be electricity
Carrier signal and telecommunication signal.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, is not described in detail or remembers in some embodiment
The part of load may refer to the associated description of other embodiments.
The steps in the embodiment of the present invention can be sequentially adjusted, merged and deleted according to actual needs.
Module or unit in system of the embodiment of the present invention can be combined, divided and deleted according to actual needs.
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
It is implemented in hardware or software, the specific application and design constraint depending on technical solution.Professional technician
Each specific application can be used different methods to achieve the described function, but this realization is it is not considered that exceed
The scope of the present invention.
In embodiment provided by the present invention, it should be understood that disclosed device/terminal device and method, it can be with
It realizes by another way.For example, device described above/terminal device embodiment is only schematical, for example, institute
The division of module or unit is stated, only a kind of logical function partition, there may be another division manner in actual implementation, such as
Multiple units or components can be combined or can be integrated into another system, or some features can be ignored or not executed.Separately
A bit, shown or discussed mutual coupling or direct-coupling or communication connection can be through some interfaces, device
Or the INDIRECT COUPLING or communication connection of unit, it can be electrical property, mechanical or other forms.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations；Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features；And these are modified
Or replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution should all
It is included within protection scope of the present invention.
Claims (10)
1. a kind of recognition methods of gait characterized by comprising
Obtain walking sample data and walking data to be identified；
According to preset model parameter, standard hidden Markov model is initialized；
Based on the walking sample data, the standard hidden Markov model after initialization is trained；
Based on viterbi algorithm, using after training standard hidden Markov model and the walking data, determine that gait is divided
Path, the gait split path include several gait spliting nodes；
According to the gait spliting node, the walking data are divided into several walking data acquisition systems, and determine each step
Line data set closes corresponding gait.
2. the recognition methods of gait as described in claim 1, which is characterized in that it is described obtain walking data to be identified it
Afterwards, comprising:
Obtain sliding window size；
Based on the sliding window size, moving average filter processing is carried out to the walking data.
3. the recognition methods of gait as described in claim 1, which is characterized in that the walking sample data includes at least one
Walking sample sequence；
It is described to be based on the walking sample data, the standard hidden Markov model after initialization is trained, comprising:
Based on forward-backward algorithm algorithm, the corresponding sequence probability of each walking sample sequence is calculated；
Judge whether the corresponding sequence probability of each walking sample sequence restrains；
Sequence probability is not restrained if it exists, then EM algorithm is based on, to the standard hidden Markov mould after the initialization
The corresponding model parameter of type carries out revaluation iteration；
If whole sequence probabilities are restrained, stop to the standard hidden Markov model training after the initialization.
4. the recognition methods of gait as described in claim 1, which is characterized in that it is described obtain walking data to be identified it
Before, comprising:
The frequency adjustment instruction comprising default frequency acquisition is sent to default Inertial Measurement Unit, the frequency adjustment instruction is used for
Indicate that the default Inertial Measurement Unit acquires user's walking data according to the default frequency acquisition；
User's walking data that the default Inertial Measurement Unit is sent are received, the walking data to be identified are obtained.
5. the recognition methods of gait as described in claim 1, which is characterized in that the gait includes heel strike to full vola
It contacts to earth, to contact to earth that liftoff to heel, heel is liftoff liftoff to toe and toe is liftoff to heel strike for full vola.
6. a kind of identifying system of gait characterized by comprising
Walking data acquisition module, for obtaining walking sample data and walking data to be identified；
Model initialization module, for being initialized to standard hidden Markov model according to preset model parameter；
Model training module carries out the standard hidden Markov model after initialization for being based on the walking sample data
Training；
Split path generation module utilizes standard hidden Markov model after training and described for being based on viterbi algorithm
Walking data determine that gait split path, the gait split path include several gait spliting nodes；
Gait divides module, for according to the gait spliting node, the walking data to be divided into several walking data
Set, and determine the corresponding gait of each walking data acquisition system.
7. the identifying system of gait as claimed in claim 6, which is characterized in that the walking data acquisition module, comprising:
Window size acquiring unit, for obtaining sliding window size；
Filter processing unit carries out moving average filter processing to the walking data for being based on the sliding window size.
8. the identifying system of gait as claimed in claim 6, which is characterized in that the walking sample data includes at least one
Walking sample sequence；
The model training module includes:
Sequence probability computing unit calculates the corresponding sequence probability of each walking sample sequence for being based on forward-backward algorithm algorithm；
Judging unit is restrained, for judging whether the corresponding sequence probability of each walking sample sequence restrains；
First processing units are not restrained for sequence probability if it exists, then EM algorithm are based on, after the initialization
The corresponding model parameter of standard hidden Markov model carries out revaluation iteration；
The second processing unit stops if restraining for whole sequence probabilities to the hidden Ma Er of standard after the initialization
It can husband's model training.
9. a kind of terminal device, including memory, processor and storage are in the memory and can be on the processor
The computer program of operation, which is characterized in that the processor realizes such as claim 1 to 5 when executing the computer program
The step of recognition methods of described in any item gaits.
10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In realization is such as the recognition methods of gait described in any one of claim 1 to 5 when the computer program is executed by processor
Step.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811407341.5A CN109492703A (en) | 2018-11-23 | 2018-11-23 | A kind of recognition methods of gait, system and terminal device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811407341.5A CN109492703A (en) | 2018-11-23 | 2018-11-23 | A kind of recognition methods of gait, system and terminal device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109492703A true CN109492703A (en) | 2019-03-19 |
Family
ID=65697618
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811407341.5A Pending CN109492703A (en) | 2018-11-23 | 2018-11-23 | A kind of recognition methods of gait, system and terminal device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109492703A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110738167A (en) * | 2019-10-14 | 2020-01-31 | 武汉大学 | pedestrian identification method based on multi-domain spatial attribute correlation analysis |
| CN110796040A (en) * | 2019-10-15 | 2020-02-14 | 武汉大学 | Pedestrian identity recognition method based on multivariate spatial trajectory correlation |
| CN111046848A (en) * | 2019-12-30 | 2020-04-21 | 广东省实验动物监测所 | Gait monitoring method and system based on animal running platform |
| CN111329485A (en) * | 2020-02-25 | 2020-06-26 | 北京机械设备研究所 | Gait recognition method and device based on IMU |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102426645A (en) * | 2011-08-30 | 2012-04-25 | 北京航空航天大学 | Multi-view and multi-state gait recognition method |
| CN107688828A (en) * | 2017-08-23 | 2018-02-13 | 武汉大学 | A kind of bus degree of crowding estimating and measuring method based on mobile phone sensor |
| CN108416325A (en) * | 2018-03-27 | 2018-08-17 | 中国计量大学 | A kind of gait recognition method of combination visual angle effect model and hidden Markov model |
-
2018
- 2018-11-23 CN CN201811407341.5A patent/CN109492703A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102426645A (en) * | 2011-08-30 | 2012-04-25 | 北京航空航天大学 | Multi-view and multi-state gait recognition method |
| CN107688828A (en) * | 2017-08-23 | 2018-02-13 | 武汉大学 | A kind of bus degree of crowding estimating and measuring method based on mobile phone sensor |
| CN108416325A (en) * | 2018-03-27 | 2018-08-17 | 中国计量大学 | A kind of gait recognition method of combination visual angle effect model and hidden Markov model |
Non-Patent Citations (1)
| Title |
|---|
| 张向刚 等: "一种基于隐马尔科夫模型的步态识别算法", 《计算机科学》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110738167A (en) * | 2019-10-14 | 2020-01-31 | 武汉大学 | pedestrian identification method based on multi-domain spatial attribute correlation analysis |
| CN110738167B (en) * | 2019-10-14 | 2022-07-19 | 武汉大学 | Pedestrian identification method based on multi-domain spatial attribute correlation analysis |
| CN110796040A (en) * | 2019-10-15 | 2020-02-14 | 武汉大学 | Pedestrian identity recognition method based on multivariate spatial trajectory correlation |
| CN110796040B (en) * | 2019-10-15 | 2022-07-05 | 武汉大学 | Pedestrian identity recognition method based on multivariate spatial trajectory correlation |
| CN111046848A (en) * | 2019-12-30 | 2020-04-21 | 广东省实验动物监测所 | Gait monitoring method and system based on animal running platform |
| CN111329485A (en) * | 2020-02-25 | 2020-06-26 | 北京机械设备研究所 | Gait recognition method and device based on IMU |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109492703A (en) | A kind of recognition methods of gait, system and terminal device | |
| EP4350572A1 (en) | Method, apparatus and system for generating neural network model, devices, medium and program product | |
| CN109241740A (en) | Malware benchmark test set creation method and device | |
| CN111143178B (en) | User behavior analysis method, device and equipment | |
| CN113535399B (en) | NFV resource scheduling method, device and system | |
| CN116032663B (en) | Privacy data processing system, method, equipment and medium based on edge equipment | |
| CN104679818A (en) | Video keyframe extracting method and video keyframe extracting system | |
| CN113746696A (en) | Network flow prediction method, equipment, storage medium and device | |
| CN113726545B (en) | Network traffic generation method and device for generating countermeasure network based on knowledge enhancement | |
| Su et al. | An iterative decision-making scheme for markov decision processes and its application to self-adaptive systems | |
| CN107608781A (en) | A kind of load predicting method, device and network element | |
| CN113312058B (en) | Similarity analysis method for intelligent contract binary function | |
| CN110796591A (en) | GPU card using method and related equipment | |
| CN114358250A (en) | Data processing method, data processing apparatus, computer device, medium, and program product | |
| CN114169506A (en) | Deep learning edge computing system framework based on industrial Internet of things platform | |
| Stylianou | Stay-point identification as curve extrema | |
| CN110533054B (en) | Multi-mode self-adaptive machine learning method and device | |
| CN112764791B (en) | Incremental update malicious software detection method and system | |
| CN112667334B (en) | Configuration method and device of equipment control information, electronic equipment and storage medium | |
| CN114595764A (en) | Method and system for acquiring influence degree of urban factors on inland inundation disaster loss | |
| CN115240843A (en) | Fairness prediction system based on structure causal model | |
| CN110781878B (en) | Target area determination method and device, storage medium and electronic device | |
| CN114331349A (en) | Scientific research project management method and system based on Internet of things technology | |
| KR20210115832A (en) | Method and apparatus for learning predictive model and apparatus for predicting time series data | |
| CN112288867A (en) | Smart building three-dimensional model rendering method and smart building system |
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: 20190319 |