EMG Processing Based on Hidden Markov Model and Support Vector Machine

摘要

This paper represents an ongoing investigation of voluntary and natural control of lower limb prosthesis using the myoelectric signals. First of all, myoelectric signals are extracted the motion features. And then these features are sent into five hidden Markov model classifiers to classify originally. In the end, based on the most possible two results of classification from hidden Markov model, the corresponding support vector machine classifiers are selected to make final decision. The results suggest that it can generalize with higher recognition rate, insensitivity to overtraining and consistent outputs demonstrating higher reliability.