Abstract
One of the most important applications of brain-computer interface (BCI) is assisting disabled people to control an external device by using motor imagery (MI). We focused in this paper on the classification of two types of MI tasks (left-hand x right-hand and right-hand x foot) in the electroencephalogram (EEG) signal. We compare various feature extraction techniques by combining common spatial pattern (CSP) with several features: variance, energy, entropy and logarithmic band power (LBP). Three types of classifiers were employed for classification: linear discriminant analysis (LDA), support vector machines (SVM) and artificial neural network (ANN). We tested our proposed method using data recorded from 17 subjects, provided by BCI-Competition III and IV. The results show that features extracted using a combination of CSP and LBP produce highest classification accuracy. LDA is more suitable than other classifiers to classify features extracted using CSP and LBP.