Personal identification by recognition of EEG power spectrogram having short processing time

Abstract

Electroencephalography (EEG) is the electrical measurement of brain activity. EEG is an interesting signal in biometrics on account of individuality and arduous to imitate the signal patterns. This research concentrates on person identification using power spectrogram without composite EEG cleaning. The underlying technique encompasses feature extraction by using spectrogram to determine dominant values from spectrogram with Singular Decomposition Value (SVD). The proposed method can separate a person from an outsider group yielding the highest accuracy at 100 percent classification in less than 1 second. The outcomes also reveal that the frontal area is an important location of scalp electrode placement in person identification.