Research On Recognition Of Sports Healthy Action Based On EMG

Modern people pay more and more attention to their own health.therefore,most people in their spare time,will invest more and more time in sports fitness.Sports injuries may occur in the process of exercise,but in the process of rehabilitation training,it is possible that the effect of rehabilitation exercise cannot be reached because the range of rehabilitation movement is not standard.Therefore,the accuracy of the movements is crucial to the recovery of health.In this paper,the author studies the methods of noise reduction in Electromyography Signal(EMG),the recognition of healthy movement,and the classification of standard/non-standard of healthy movement.1.EMG signal is weak and extremely vulnerable to interference.Removing the noise contained in EMG signals and improving the signal purity will greatly improve the accuracy of sports health movement classification.In this paper,an innovative noise reduction method of EMG signal-software and hardware combined noise reduction method is proposed.Firstly,the hardware noise reduction method is used to reduce static interference,noise introduced due to physiological characteristics,and environmental noise.Then the instrument’s inherent noise is removed by software noise reduction method.The experimental results show that the noise reduction method combined with software and hardware proposed in this paper can effectively remove the noise in EMG.2.Research on EMG-based sports health activity recognition firstly calculates seven statistics of EMG as the characteristics of EMG signal,including EMG integral EMG value,absolute value integral,maximum value,mean value,minimum value,root mean square and variance.This experiment maximizes the extraction of features from the original EMG data for the use of algorithms and models.Seven kinds of EMG signal based on the above features,this article uses the Support Vector machine(Support Vector Mechine,SVM),Random forests,Gridient Boosting algorithm,such as to bend to sports such as running,jumping,and standing up health action.The experiment shows that the research accuracy of EMG-based motion health classification proposed in this paper has high classification accuracy3.When patients make healthy movement,they do not make standard healthy movement because they do not have comprehensive cognition of some targeted movements.In this paper,an innovative action standard/non-standard judgment method is proposed: first,EMG signals are segmented and the spectrum of each EMG signal is calculated,and then the spectrum is input into the convolutional neural network established in this paper,and the output of the convolutional neural network is taken as the input of the decision tree to determine the action standard/non-standard.To avoid prolonged convalescence due to improper movement of healthy sports.