Comparison Analysis of Overt and Covert Mental Stimuli of Brain Signal for Person Identification

摘要

Cybersecurity is an important and challenging issue faced by the governments, financial institutions and ordinary citizens alike. Secure identification is needed for accessing confidential personal information, online bank transactions, people's social networks etc. Brain signal electroencephalogram (EEG) can play a vital role in ensuring security as it is non-vulnerable and very difficult to forge. In this article, we develop an EEG based biometric security system. The purpose of this research is to find the relationship between thinking capability and person identification accuracy by comparison analyzing of overt and covert mental stimuli of brain signal. The Discrete Wavelet Transform (DWT) is used to extract different significant features which separate Alpha, Beta and Theta band of frequencies of the EEG signal. Extracted EEG features of different bands and their combinations such as alpha-beta, alpha-theta, theta-beta, alpha-beta-theta are classified using an artificial neural network (ANN) trained with the back propagation (BP) algorithm. Another classifier K-nearest neighbors (KNN) is used to verify the results of this experiment. Both classification results show that alpha band has a higher convergence rate than other bands, beta and theta, for the overt EEG signal. From overt mental stimuli, we also discover that individual band provides better performance than band combination. So, we have applied Back Propagation (BP) algorithm at individual band of various features of covert mental stimuli and obtained the accuracy 73.1%, 78.1% and 74.4% for alpha, beta and theta band respectively. By comparing the analysis of overt and covert mental stimuli, the overt brain signal shows better performance. Finally, we conclude that the relationship between thinking capability and person identification accuracy is inversely proportional. The results of this study are expected to be helpful for future research by using various thinking capability brain signals based biometric approaches.