Research On Driving Fatigue Detection With Fusion Of EEG And Forehead EOG

There are various studies on driving fatigue detection with physiological signals.Among these signals,electroencephalogram(EEG)is considered to be the most direct,effective,and promising one to detect driving fatigue.Electrooculogram(EOG)is another effective signal for driving fatigue detection.Different from traditional EOGs which are collected from the places around the eyes,forehead EOGs are collected form the forehead of the subjects,which will not affect the normal activities of the subjects,but also conducive to the integration of wearable devices and practical applications.Although EEG and forehead EOG both can be used to detect driving fatigue,EEG or forehead EOG only focuses on a certain specific aspect and the driver's fatigue level only can be inferred from the information available,it is difficult to build a roubst fatigue detection model by using single modality.This paper used different methods to fuse EEG and forehaed EOG,and then different machine learning algorithms were used to train detection models.We collected EEG data from the key brain areas related to fatigue.Two fusion strategies were adopted: feature level fusion(FLF)and decision level fusion(DLF).At feature level,the EEG feature vector and forehead EOG feature vector are directly concentrated into a larger feature vector,which will be used as the input to train model.At decision level,the mean of the two regression outcomes will be computed as the final estimated fatigue level.PERCLOS(percentage of eye closure)is calculated by using the eye movement data recorded by eye tracking glasses as the indicator of drivers' fatigue level.The prediction correlation coefficient and root mean square error(RMSE)between the estimated fatigue level and the real fatigue level are both used to evaluate the performance of single modality and fusion modality.A comparative study on modality performance is conducted between discriminative graph regularized extreme learning machine(GELM)and support vector machine(SVM).The experimental results show that fusion modality can improve the performance of driving fatigue detection with a higher prediction correlation coefficient and a lower RMSE value in comparison with solely using EEG or forehead EOG.And FLF achieves better performance than DLF.GELM is more suitable for driving fatigue detection than SVM.