Feature extraction and classification of steady-state visual evoked potential in application for brain computer interface systems

Brain Computer Interface (BCI) system improves people ability to send massages and comments without movement, in the same time BCI system asists disable people to contact with their environment by removing motor function limitation. In this study, different signal processing methods are applied over steadt-state visula evoked potential (SSVEP) signals, induced from brain electrical activity. Filtering, feature extraction, classification are main steps of signal processing. For fitlerin; Butterworth fitler, applied fitler, LMS algorithma are investigated. TF- based fiature extraction and Single Value distribution ( SVD) are used for feature extraction. For classification K En close ajusent, wavelet transform paket and vector machins methods are consentrated. The result of these methods is investigated over aqcuired real data from lablatory of technical university of Graz. The results are compared to each other and significant information is obtained.