On-line EEG classification for brain-computer interface based on CSP and SVM

Brain-Computer Interface (BCI) research aims at automatically translating neural commands into control signals through classifying the electroencephalogram (EEG) patterns of different mental tasks (e.g. imagined hand and foot movements). This paper presents a method of on-line classification for BCI based on Common Spatial Pattern (CSP) for feature extraction and Support Vector Machine (SVM) as a classifier. The best classification results for three subjects are 86.3%, 91.8%, and 92%. The high classification rate in a real-time 3D computer game indicates that the proposed method is promising for an EEG-based brain-computer interface. It can provide a new way for the EEG automation classification when the EEG is used an input signal to a brain computer interface.