Analysis and Classification of EMG Signal Using Labview With Different Weights

Abstract

Prosthetic is a branch of biomedical engineering that deals with missing human body parts with artificial one. The present research work is broadly divided in two parts namely; detection and classification of EMG signals and replacement of missing body parts. Further, the work carried out is related to feature extraction and classification of EMG signals obtained from bicep and below elbow position. The EMG signals are obtained for different cases: Grasping and Lifting, with different weight combinations. The SEMG signal is obtained from the surface of the body by using disposable electrodes. For the acquisition of SEMG signal BIOKIT Datascope system is used. The features are extracted from the conditioned EMG signal such as: root mean square value, integrated EMG value, mean absolute value and zero crossing rate. The signals are usually non-repeatable and contradictory in nature. Therefore, to classify such kind of signal, a classifier able to withstand uncertainties in data is required. Fuzzy theory is well known for its capability to deal with imprecise environment. So, in this work a fuzzy classifier is designed and implemented using LabVIEW software. The classifier system is tested using 30 subjects. The simulation results have authenticated the capability of implemented system.