Overview of EMG Signal Preprocessing and Classification for Bionic Hand Control

This paper deals with the field of Biological Signal especially one of the most used signals in the field of prosthetic devices, which is the EMG signal. After introducing this kind of signal, we will talk about different techniques used for preprocessing it in order to prepare it for classification tasks. We Will then overview different noises that affect those sensible type of signals, different techniques applied for filtering them and the useful envelope detection methods that can be applied on EMG. The next part of this paper will look at some famous feature extraction and classification techniques that proved themselves useful when dealing with classification tasks of biological signals. Finally, we will present the mechanical design of a multiarticulate multifunctional prosthetic hand, the promising classification results that can gives 98.3% and our perspectives.