Feature extraction for identification of extension and flexion movement of wrist using EMG signals

Electromyography (EMG) is an experimental technique developed for the purpose of studying muscle movement. Information from raw EMG signals is extracted by application of “Wavelet Transform”. This technique holds good in handling non stationary signals which ordinary Fourier Transform and even Short Time Fourier Transform fail to handle. In this work, by applying wavelet transform, signal was first de-noised and then some unique parameters were calculated. This set of features were then used as reference to identify different movements. Later, signals from test subjects were acquired and the same features were extracted. A cost function is used to identify the movement. This research provides a base for prosthetic arm designing.