Locomotion classification using EMG signal

Abstract

This work gives a comparative study on the use of Linear Discriminant Analysis (LDA), Artificial Neural Network (ANN) and Naive-Bayes Classifier (NBC) for recognizing various locomotion modes using parameters derived from the transient EMG signals taken from healthy subjects and thus provide a better control mechanism for lower limb prosthesis. These classifiers have been taken into consideration owing to their extensive use in various real-time applications.