Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A method for feature extraction of electromyographic signals

An electromyographic signal and feature extraction technology, applied in neural learning methods, biometric recognition, instruments, etc., can solve the problems of insufficient frequency domain feature extraction and large time domain feature changes, and reduce input data bandwidth and redundancy. In addition, to solve the effect of large changes in time domain features

Active Publication Date: 2022-07-15
HOHAI UNIV CHANGZHOU
View PDF16 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to solve the above problems, the present invention proposes a method for feature extraction of electromyographic signals to realize feature extraction of electromyographic signals and solve the problems of large changes in time domain features and insufficient frequency domain feature extraction faced by existing electromyographic signal extraction methods

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A method for feature extraction of electromyographic signals
  • A method for feature extraction of electromyographic signals
  • A method for feature extraction of electromyographic signals

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] The technical solutions of the present invention will be further elaborated below according to the accompanying drawings and in conjunction with the embodiments.

[0031] The EMG signal data used in this embodiment is the EMG signal data obtained by four people performing ten different gestures. The different gestures of each person represent a category, and each gesture is an 8-dimensional data. The data has 10000 sample points. An electromyographic signal feature extraction method, constructing a stacked autoencoder network architecture, that is, stacking multiple autoencoders, each autoencoder is a three-layer neural network, setting the parameters of each layer of neural network and using stochastic gradients Stochastic gradient descent (SGD) is used to train and update weights; a restricted Boltzmann machine (RBM) with only visible and hidden layers is established, and gradient descent is used to train and update weights. After training the autoencoder and the res...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an electromyographic signal feature extraction method. First, the electromyographic signal is subjected to Fourier transform to obtain Fourier coefficients, and the Fourier coefficients of a preset frequency band are selected and normalized; a stacked automatic encoder is constructed, and a set of The network parameters of each layer are trained to update the parameters; the stacked restricted Boltzmann machine is constructed, and the network parameters of each layer are set to train and update the parameters; after the pre-trained autoencoder and the restricted Boltzmann machine are completed, the The two parts of pre-trained parameters are connected before and after, that is, the output of the auto-encoder is the input of the restricted Boltzmann machine, forming a through network to realize the feature extraction of EMG signals. The time-domain features vary greatly and the frequency-domain feature extraction is insufficient.

Description

technical field [0001] The invention relates to the technical field of signal feature extraction, in particular to an electromyographic signal feature extraction method. Background technique [0002] EMG signals can reflect neuromuscular activity to a certain extent, and have important practical value in clinical medicine, ergonomics, rehabilitation medicine, and sports science. Feature extraction can discover more meaningful latent variables, help to gain a deeper understanding of EMG data, reduce data storage and input data bandwidth, and reduce redundancy. [0003] The traditional method of EMG signal processing is to regard EMG signal as a random signal whose mean value is zero and whose variance varies with the change of signal strength. The extraction of time-domain features is relatively simple, so time-domain analysis methods have been widely used in the field of EMG signal applications. However, although the time-domain characteristics are easy to extract, a large...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/08
CPCG06N3/084G06V40/10G06F18/214
Inventor 徐宁刘妍妍倪亚南潘安顺刘小峰
Owner HOHAI UNIV CHANGZHOU
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com