Steady-state Visual Fatigue Detection Based On Hilbert-Huang Transform

With the pace of work and life gradually accelerates,fatigue has become a common state of mind.In the fields of traffic driving,aerospace,human-machine system monitoring,etc.,the inattention,slow response,and mental distraction caused by the fatigue of the brain are likely to cause serious accidents and cause huge property losses Therefore,monitoring the state of brain fatigue is of great significance.The study found that EEG signals can reflect the mental state of the human body.Therefore,it is possible to predict brain fatigue by detecting the characteristics of the EEG signal under fatigueSteady-state visual evoked potential means that when a human cerebral cortex is subjected to a fixed-frequency visual stimulus,it produces a continuous response related to the stimulation frequency.Steady-state visual evoked potentials are widely used in the study of EEG signals because of their simple signal acquisition and no training.This article also uses steady-state visual evoked potentials to induce fatigue electroencephalography in subjects.The main work of the article is as followsI.The experiment designed a stimulus source of black and white chess.The excitation source's flicker frequency is between 9-13Hz.Subjects were placed in a quiet,well-lit experimental environment to receive long-term visual stimuli that produced visual fatigue EEG signals.During the course of the experiment,subjects completed the subjective fatigue scale as a reference for subjective statistics as required.The independent component analysis method was used to remove artifacts such as ocular electricity and myoelectricity in the original EEG signal2.Using the nonlinear non-stationary signal processing method Hilbert-Huang Tansform(HHT)to extract the frequency domain features in EEG signals.The EMF component is obtained by empirical mode decomposition of the EEG signal,and the IMF component with the original signal correlation coefficient greater than 50%is selected according to the Pearson coefficient for reconstruction.The Hilbert transform is performed on the reconstructed signal,and the Hilbert marginal spectral energy of the reconstructed signal is calculated.It is found that the marginal energy spectrum amplitudes of the four EEG brain electrical rhythms ?,?,(?+?)/?,?/? in the EEG signal are in the awake state and fatigue.There is a significant difference in the state.The Hilbert marginal spectral amplitudes of the four EEG rhythms are extracted as feature vectors.3.Comparing the advantages and disadvantages of several classification algorithms,the article decided to use Support Vector Machine(SVM)as the classifier of EEG signals.The feature vector is transmitted to the SVM.The classification results show that:?,?/? and the combined feature vectors of the four EEG indicators T can be well differentiated,which can be used as an indicator to judge the fatigue state.