A new multi-level approach to EEG based human authentication using eye blinking

摘要: A new multi-level EEG biometric authentication system based on eye blinking EOG signals is proposed.Eye features time delineation.Two approaches for the proposed feature and score level fusion.Multi-level achieved higher recognition rates than single one using only. This letter proposes a approach human Electro-Encephalo-Gram (EEG) (brain waves) Electro-Oculo-Gram (EOG) signals. The main objective of this to improve performance which are considered as source artifacts EEG. Feature fusion tested system. Density canonical correlation analysis strategies applied fusions, respectively. Autoregressive modeling (during relaxation or visual stimulation) delineation waveform adopted extraction stage. Finally, classification stage performed linear discriminxant analysis. For evaluation, database 31 subjects performing three different tasks relaxation, stimulation, was collected Neursky Mindwave headset. Using features, significant improvement achieved, in terms correct equal error rates, over