A Novel Method of De-noising and Classifying on Mental EEG of Imaging Left-Right Hands Movement

摘要

After analyzing the current wavelet threshold de-noising methods and independent component analysis (ICA) methods in EEG, this paper proposed a novel method for EEG de-noising which combines the new threshold de-noising method with ICA method and implicates it to deal with mental EEG of imaging left-right hands movement, and then classifies the signal by Support Vector Machine (SVM). The correct classification rate of 89.93% is achieved by the approach in this paper.