To further improve the accuracy and speed of gesture recognition in the field of human-computer interaction,and explore the influence of muscle fatigue on gesture recognition,an improved GA-BLS method was proposed,genetic algorithms(GA)were used to optimize the parameters of the broad learning system(BLS)model,and elastic network regression was used to improve the traditional BLS model.The proposed model was used to analyze the A-mode ultrasound signal and EMG signal under eight kinds of gestures for gesture recognition,and compared with SVM,KNN,RF,LDA and other methods to verify the effectiveness of the research methods.Furthermore,the A-mode ultrasound and EMG in a long period of time were divided into four data segments.It was found that with the increase of muscle fatigue,the accuracy of gesture recognition showed a significant downward trend,and A-mode ultrasound signal had better fatigue resistance than EMG signal.
维普
万方数据
