CN109567798A - Daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines - Google Patents

Daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines Download PDF

Info

Publication number
CN109567798A
CN109567798A CN201811603106.5A CN201811603106A CN109567798A CN 109567798 A CN109567798 A CN 109567798A CN 201811603106 A CN201811603106 A CN 201811603106A CN 109567798 A CN109567798 A CN 109567798A
Authority
CN
China
Prior art keywords
support vector
coherence
vector machines
wavelet
electromyography signal
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
Application number
CN201811603106.5A
Other languages
Chinese (zh)
Inventor
席旭刚
杨晨
石鹏
袁长敏
罗志增
张启忠
佘青山
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN201811603106.5A priority Critical patent/CN109567798A/en
Publication of CN109567798A publication Critical patent/CN109567798A/en
Pending legal-status Critical Current

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/389—Electromyography [EMG]
    • 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
    • A61B5/1116—Determining posture transitions
    • 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
    • A61B5/1116—Determining posture transitions
    • A61B5/1117—Fall detection
    • 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
    • A61B5/1123—Discriminating type of movement, e.g. walking or running
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Biophysics (AREA)
  • General Health & Medical Sciences (AREA)
  • Veterinary Medicine (AREA)
  • Public Health (AREA)
  • Animal Behavior & Ethology (AREA)
  • Surgery (AREA)
  • Molecular Biology (AREA)
  • Medical Informatics (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Pathology (AREA)
  • Biomedical Technology (AREA)
  • Physiology (AREA)
  • Artificial Intelligence (AREA)
  • Signal Processing (AREA)
  • Psychiatry (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Dentistry (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Fuzzy Systems (AREA)
  • Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)

Abstract

The daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines that the invention discloses a kind of, the present invention acquires the electromyography signal of human body related muscles by electromyographic signal collection instrument, the sample data for obtaining two-way electromyography signal, is pre-processed using a kind of improvement Threshold Denoising method.Calculate the wavelet coherence of two-way electromyography signal.Classification and Identification is carried out using obtained wavelet coherence as feature vector input support vector machines, successfully identifies different daily behaviors, discrimination with higher.Method of the present invention by the myoelectricity feature of small echo coherence in conjunction with support vector machines identifies discrimination with higher and reliability to human body daily behavior.The experimental results showed that the method for the present invention to upstairs, downstairs, stand, walking, running, the average sensitivity fallen up to 96.17, average specificity is up to 92.29, higher than general traditional method.

Description

Daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines
Technical field
The invention belongs to area of pattern recognition, are related to a kind of Method of Surface EMG Pattern Recognition, in particular to one kind is based on The daily behavior recognition methods of myoelectricity small echo coherence and support vector machines.
Background technique
It is had been to be concerned by more and more people at present about the identification of daily behavior, this is the base of behavior monitoring and fall detection Plinth.Wearable sensor is that have easy to operate, the strong feature of environment compatibility using most wide method in behavioral value. Wearable sensor includes movement and electro-physiological sensor, such as accelerator sensor, inertial sensor, gyroscope and surface flesh Electric signal (surface electromyography, sEMG) sensor.Wherein sEMG sensor can directly reflect that human body is each The movable electrophysiologic response of kind, the calculating time is shorter, can distinguish passive behavior and active behavior, is widely used to behavior knowledge Not, gait analysis and prosthesis control field.However, being still had in terms of finding best features collection from original sEMG data set Sizable challenge.
Wavelet transformation has a wide range of applications in surface electromyogram signal processing, it is by Decomposition Surface EMG at many Subband comprising precise information.And coherent analysis, and number are carried out to the relationship in time frequency space between two kinds of echo signals Common method in word signal analysis especially wavelet transformation.Ryotaro Imoto et al. is with relevant point of electromyography signal small echo Analysis has studied the coordinated movement of various economic factors mechanism of agonistic muscle and Opposing muscle, has obtained under stable condition there is more high correlation than instability condition Conclusion.In general, small echo coherence can be used to analyze nonstationary random signal, such as electromyography signal and EEG signals.Surface flesh The coherence of electric signal can provide a seed coat layer muscle coupling information.Currently based on the day of surface electromyogram signal coherent analysis Normal Activity recognition research is less, there is more wide research space.
Sorting technique is another major issue of behavior monitoring.Support vector machines is the human body based on surface electromyogram signal An important application in classification of motion field, support vector machines construct one or a set of super flat in higher-dimension or infinite dimensional space Face, to separate different data sets.Compared with traditional neural network, support vector machines provides a kind of phase for optimization problem To simple method for solving, suitable for the high dimensional feature vector categorizing system based on sEMG.
Summary of the invention
It is effectively based on myoelectricity daily behavior recognition methods in order to invent a kind of stabilization, is proposed a kind of based on myoelectricity small echo phase The daily behavior recognition methods of stemness and support vector machines.The myoelectricity letter of human body related muscles is acquired by electromyographic signal collection instrument Number, the sample data of two-way electromyography signal is obtained, is pre-processed using a kind of improvement Threshold Denoising method.Calculate two-way The wavelet coherence of electromyography signal.Divided obtained wavelet coherence as feature vector input support vector machines Class identification, successfully identifies different daily behaviors, discrimination with higher.
In order to achieve the goal above, the method for the present invention mainly comprises the steps that
Step 1 acquires the electromyography signal of human body related muscles by electromyographic signal collection instrument, obtains two-way electromyography signal x (t) it with the sample data of y (t), is then pre-processed using a kind of improvement Threshold Denoising method.
diWithWavelet coefficient respectively before and after threshold process, λ are threshold value, and N is normal number.
Step 2 calculates the wavelet coherence of two-way electromyography signal.
Wx(a, b) is wavelet coefficient, It is that mother Morlet is small Wave function, a are scale factors, and b is time shift, and t is local time's origin, Wxy(a, b) is cross wavelet analysis,
Step 3 is inputted support vector machines using the obtained wavelet coherence of step 2 as feature vector and classified Identification.The minimum value that support vector machines needs solution to find following equation carrys out optimization problem:
The formula is obeyed:
yi[(xi·ω)+b]≥1-ξi, i=1,2 ..., N
Wherein ω is the normal vector of hyperplane, ξiIt is the minimum nonnegative number for meeting above-mentioned equation, x is n-dimensional vector, and b is one A scalar, c are independent variable, and N is sample size, and y is the model to be learnt.
Following equation is defined using kernel function:
Wherein N indicates the quantity of supporting vector, k (x, xi) it is kernel function, aiIt is to be by what solution optimization problem obtained Number.
The wavelet coherence based on electromyography signal and support vector machines daily behavior recognition methods that the present invention designs, tool There are following features:
Method of the present invention by the myoelectricity feature of small echo coherence in conjunction with support vector machines identifies human body daily behavior Discrimination with higher and reliability.The experimental results showed that the method for the present invention to upstairs, go downstairs, stand, walk, run, fall Average sensitivity is up to 96.17, and average specificity is up to 92.29, higher than general traditional method.
Detailed description of the invention
Fig. 1 is implementation flow chart of the invention；
Fig. 2 (a) is the electromyography signal figure of related muscles under standing behavior；
Fig. 2 (b) is the electromyography signal figure of related muscles under walking behavior；
Fig. 2 (c) is the electromyography signal figure of related muscles under running behavior；
Fig. 2 (d) is the electromyography signal figure of related muscles under daily behavior upstairs；
Fig. 2 (e) is the electromyography signal figure of related muscles under going downstairs daily behavior；
Fig. 2 (f) is the electromyography signal figure of related muscles under tumble daily behavior；
Fig. 3 is each the 32nd multi-scale wavelet coherence factor histogram of muscle combination under different daily behaviors；
Fig. 4 is the wavelet coherence scatter plot of 6 kinds of different muscle combinations in different daily behaviors.
Specific embodiment
As shown in Figure 1, the present embodiment includes the following steps:
Step 1 obtains the sample data of two-way electromyography signal x (t) and y (t), acquires people by electromyographic signal collection instrument The electromyography signal of body related muscles, specifically: being adopted by DELSYS Trigno Wireless System electromyographic signal collection instrument The electromyography signal of related muscles when collecting human body lower limbs movement, the experiment movement taken are to stand, walk, running, going upstairs, downstairs Ladder and tumble, the related muscles taken are gastrocnemius, tibialis anterior, rectus femoris and semitendinosus, and Fig. 2 (a)-(f) is different daily The electromyography signal figure of related muscles under behavior.Then it is pre-processed using a kind of improvement Threshold Denoising method.
WhereindiWithWavelet coefficient respectively before and after threshold process, N are normal numbers.
Step 2 calculates the wavelet coherence of two-way electromyography signal.
Wx(a, b) is wavelet coefficient, It is that mother Morlet is small Wave function, a are scale factors, and b is time shift, and t is local time's origin, Wxy(a, b) is cross wavelet analysis,Fig. 3 show each the 32nd multi-scale wavelet coherence factor of muscle combination under different daily behaviors Histogram；Fig. 4 is the wavelet coherence scatter plot of 6 kinds of different muscle combinations in different daily behaviors.
Step 3 carries out Classification and Identification to data using support vector machines.Support vector machines, which needs to solve, finds following side The minimum value of journey carrys out optimization problem:
The formula is obeyed:
yi[(xi·ω)+b]≥1-ξi, i=1,2 ..., N
Wherein ω is the normal vector of hyperplane, ξiIt is the minimum nonnegative number for meeting above-mentioned equation, x is n-dimensional vector, and b is one A scalar, c are independent variable, and N is sample size, and y is the model to be learnt.
Following equation is defined using kernel function:
Wherein N indicates the quantity of supporting vector, k (x, xi) it is kernel function, aiIt is to be by what solution optimization problem obtained Number.
Final recognition result is as shown in table 1.
The classification results (sensitivity, specificity, %) of 16 daily behaviors of table

Claims (1)

1. the daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines, it is characterised in that: this method includes Following steps:
Step 1 acquires the electromyography signal of human body related muscles by electromyographic signal collection instrument, obtains two-way electromyography signal x (t) and y (t) then sample data is pre-processed using a kind of improvement Threshold Denoising method；
diWithWavelet coefficient respectively before and after threshold process, λ are threshold value, and N is normal number；
Step 2 calculates the wavelet coherence through the pretreated two-way electromyography signal of step 1；
Wx(a, b) is wavelet coefficient, It is Morlet mother wavelet function, a It is scale factor, b is time shift, and t is local time's origin, Wxy(a, b) is cross wavelet analysis,
Step 3 is inputted support vector machines using the obtained wavelet coherence of step 2 as feature vector and carries out classification knowledge Not；Support vector machines needs to solve the minimum value of following equation to optimize:
The formula is obeyed:
yi[(xi·ω)+b]≥1-ξi, i=1,2 ..., N
Wherein ω is the normal vector of hyperplane, ξiIt is the minimum nonnegative number for meeting above-mentioned equation, x is n-dimensional vector, and b is a mark Amount, c is independent variable, and N is sample size, and y is the model to be learnt；
Following equation is defined using kernel function:
Wherein N indicates the quantity of supporting vector, k (x, xi) it is kernel function, aiIt is the coefficient obtained by solving optimization problem.
CN201811603106.5A 2018-12-26 2018-12-26 Daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines Pending CN109567798A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811603106.5A CN109567798A (en) 2018-12-26 2018-12-26 Daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811603106.5A CN109567798A (en) 2018-12-26 2018-12-26 Daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines

Publications (1)

Publication Number Publication Date
CN109567798A true CN109567798A (en) 2019-04-05

Family

ID=65932729

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811603106.5A Pending CN109567798A (en) 2018-12-26 2018-12-26 Daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines

Country Status (1)

Country Link
CN (1) CN109567798A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110151176A (en) * 2019-04-10 2019-08-23 杭州电子科技大学 A kind of continuous method for estimating of upper limb elbow joint based on electromyography signal
CN110232976A (en) * 2019-07-01 2019-09-13 上海电机学院 A kind of Activity recognition method based on the measurement of waist shoulder surface myoelectric
WO2021042971A1 (en) * 2019-09-03 2021-03-11 北京海益同展信息科技有限公司 Surface electromyogram signal processing method and apparatus, and wearable device
CN112932508A (en) * 2021-01-29 2021-06-11 电子科技大学 Finger activity recognition system based on arm electromyography network

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101833671A (en) * 2010-03-30 2010-09-15 西安理工大学 Support vector machine-based surface electromyogram signal multi-class pattern recognition method
CN101859377A (en) * 2010-06-08 2010-10-13 杭州电子科技大学 Electromyographic signal classification method based on multi-kernel support vector machine
CN103761424A (en) * 2013-12-31 2014-04-30 杭州电子科技大学 Electromyography signal noise reducing and aliasing removing method based on second-generation wavelets and ICA (independent component analysis)
CN104107042A (en) * 2014-07-10 2014-10-22 杭州电子科技大学 Electromyographic signal gait recognition method based on particle swarm optimization and support vector machine
CN105997064A (en) * 2016-05-17 2016-10-12 成都奥特为科技有限公司 Method for identifying human lower limb surface EMG signals (electromyographic signals)
US20170258390A1 (en) * 2016-02-12 2017-09-14 Newton Howard Early Detection Of Neurodegenerative Disease
CN107817731A (en) * 2017-11-27 2018-03-20 中国兵器工业计算机应用技术研究所 Merge unmanned the platform control system and control method of myoelectricity and brain electric information

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101833671A (en) * 2010-03-30 2010-09-15 西安理工大学 Support vector machine-based surface electromyogram signal multi-class pattern recognition method
CN101859377A (en) * 2010-06-08 2010-10-13 杭州电子科技大学 Electromyographic signal classification method based on multi-kernel support vector machine
CN103761424A (en) * 2013-12-31 2014-04-30 杭州电子科技大学 Electromyography signal noise reducing and aliasing removing method based on second-generation wavelets and ICA (independent component analysis)
CN104107042A (en) * 2014-07-10 2014-10-22 杭州电子科技大学 Electromyographic signal gait recognition method based on particle swarm optimization and support vector machine
US20170258390A1 (en) * 2016-02-12 2017-09-14 Newton Howard Early Detection Of Neurodegenerative Disease
CN105997064A (en) * 2016-05-17 2016-10-12 成都奥特为科技有限公司 Method for identifying human lower limb surface EMG signals (electromyographic signals)
CN107817731A (en) * 2017-11-27 2018-03-20 中国兵器工业计算机应用技术研究所 Merge unmanned the platform control system and control method of myoelectricity and brain electric information

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
RYOTARO IMOTO等: "Preliminaly Study on Coordinated Movement Mechanism of Multiple Muscle Using Wavelet Coherence Analysis", 《2016 5TH IIAI INTERNATIONAL CONGRESS ON ADVANCED APPLIED INFORMATICS》 *
高文仲: "小波分析在图像去噪中的应用研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110151176A (en) * 2019-04-10 2019-08-23 杭州电子科技大学 A kind of continuous method for estimating of upper limb elbow joint based on electromyography signal
CN110232976A (en) * 2019-07-01 2019-09-13 上海电机学院 A kind of Activity recognition method based on the measurement of waist shoulder surface myoelectric
WO2021042971A1 (en) * 2019-09-03 2021-03-11 北京海益同展信息科技有限公司 Surface electromyogram signal processing method and apparatus, and wearable device
CN112932508A (en) * 2021-01-29 2021-06-11 电子科技大学 Finger activity recognition system based on arm electromyography network
CN112932508B (en) * 2021-01-29 2022-10-14 电子科技大学 Finger activity recognition system based on arm electromyography network

Similar Documents

Publication Publication Date Title
CN109567798A (en) Daily behavior recognition methods based on myoelectricity small echo coherence and support vector machines
Phinyomark et al. Surface electromyography (EMG) signal processing, classification, and practical considerations
Benson et al. Classifying running speed conditions using a single wearable sensor: Optimal segmentation and feature extraction methods
Burns et al. Upper limb movement classification via electromyographic signals and an enhanced probabilistic network
CN111046731B (en) Transfer learning method and recognition method for gesture recognition based on surface electromyographic signals
Ahsan et al. Hand motion detection from EMG signals by using ANN based classifier for human computer interaction
Kuncan et al. A novel approach for activity recognition with down-sampling 1D local binary pattern
CN109567799A (en) EMG Feature Extraction based on smooth small echo coherence
Oung et al. Wearable multimodal sensors for evaluation of patients with Parkinson disease
Oleinikov et al. Feature extraction and real-time recognition of hand motion intentions from EMGs via artificial neural networks
Bergil et al. Efficient hand movement detection using k-means clustering and k-nearest neighbor algorithms
Schmid et al. SVM versus MAP on accelerometer data to distinguish among locomotor activities executed at different speeds
Noor et al. Detection of freezing of gait using unsupervised convolutional denoising autoencoder
Zhang et al. Gait pattern recognition based on plantar pressure signals and acceleration signals
Luan et al. Har-semg: A dataset for human activity recognition on lower-limb semg
Wen et al. [Retracted] Multimodal Sensor Motion Intention Recognition Based on Three‐Dimensional Convolutional Neural Network Algorithm
Tomczyński et al. Hand gesture-based interface with multichannel sEMG band enabling unknown gesture discrimination
Dwiyantoro et al. A simple hierarchical activity recognition system using a gravity sensor and accelerometer on a smartphone
CN110151176A (en) A kind of continuous method for estimating of upper limb elbow joint based on electromyography signal
CN115399792A (en) Muscle identification method based on posture division
Ge et al. Feature extraction and classification of hand movements surface electromyogram signals based on multi-method integration
Ahlawat et al. DWT-based hand movement identification of EMG signals using SVM
Kołodziej et al. Registration and analysis of acceleration data to recognize physical activity
Belsare et al. Performance comparison of different EEG analysis techniques based on deep learning approaches
Peng et al. Evaluation of sEMG-based feature extraction and effective classification method for gait phase detection

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

Application publication date: 20190405

RJ01 Rejection of invention patent application after publication