Development of EMG classifier adapting to operator's characteristics : Quantification of the attainment of motor skill from frequency of data using

Abstract

In the case of classifying human's motion from electromyogram (EMG), it needs to estimate an attainment of operator's motor skill. This is because it is used to judgement of the adequate number of motion for the operator and to judgement of a necessity of the operator's training. However, previous works have not consider the attainment of motor skill. Accordingly, this paper proposes a quantification algorithm of the attainment of motor skill and appends it in our proposed EMG classifier. Experimental results reported that the attainment of motor skill enables to evaluate the operator's current motor skill. Moreover, comparing the attainment and a classification rate enables to discuss the sufficiency of information amount from the EMG signals.