Introduction to Digital Signal Processing and Filter Design
Introduction to Digital Signal Processing and Filter Design
Hi-index | 0.01 |
This paper treats the effects of EMG signal sampling frequency and the pass-band frequency on neuromuscular signal recognition which are of a great importance. It improves the optimisation of the choice of those parameter values that guarantees the best performance. The considered EMG signal sampling frequency and pass-band frequency have a significant importance in viewpoint of the extracted information from this signal. It will be shown that this information depends on both sampling and pass-band frequency values. The signal information do not have the same pass-band frequency location and also do not have the same sampling frequency value for all the features, but these parameters are then self depending on the type of the feature. In this case the classification of the EMG signals is done by using only the beginning part of the signal, which is equal in our case to 256ms. The classification is done using two intelligent computational methods: Radial Basis Function network (RBF) and Fuzzy Subtractive Clustering network (FSC). Each method has been applied using four different values of spread and radius values corresponding to RBF and FSC respectively.