EEG Decoding of Motor Imagery Mode Based on WPT-CSP

In this paper, wavelet packet transform (WPT) and common space pattern (CSP) algorithm are used to extract signal features. Firstly, the wavelet packet transform is used to decompose the signal, and the appropriate wavelet packet subband is selected for reconstruction. Then, feature vectors are extracted by common space pattern projection mapping. Finally, it innovatively combines linear discriminant classifier (LDA) to classify. The results on BCI-Competition-IV-2b data set show that the average classification accuracy is 86.29%.