Autore: Wei-Yen Hsu
Formato: Copertina flessibile
Pagine: 52
Data Pubblicazione: 2013-04-02
Edizione: 1
Lingua: English
Descrizione:
1. In this book, a novel braincomputer interface (BCI) system is proposed to analyze motor imagery (MI) electroencephalogram (EEG) signals. 2. After eliminating EOG artifacts automatically and extracting features by waveletbased phase synchronization approach, support vector machine (SVM) is adopted for the classification of singletrial left and right MI data. 3. The EOG artifacts are automatically removed by means of modified independent component analysis (ICA). 4. The features are extracted from wavelet data by phase synchronization, and then classified by the SVM. 5. Compared with the results without EOG artifact removal, spectral band and AR model features, the proposed system achieves satisfactory results in BCI applications.