EEG-based drowsiness estimation for safety driving using independent component analysis

Abstract

Preventing accidents caused by drowsiness has become a major focus of active safety driving in recent years. It requires an optimal technique to continuously detect drivers' cognitive state related to abilities in perception, recognition, and vehicle control in (near-) real-time. The major challenges in developing such a system include: 1) the lack of significant index for detecting drowsiness and 2) complicated and pervasive noise interferences in a realistic and dynamic driving environment. In this paper, we develop a drowsiness-estimation system based on electroencephalogram (EEG) by combining independent component analysis (ICA), power-spectrum analysis, correlation evaluations, and linear regression model to estimate a driver's cognitive state when he/she drives a car in a virtual reality (VR)-based dynamic simulator. The driving error is defined as deviations between the center of the vehicle and the center of the cruising lane in the lane-keeping driving task. Experimental results demonstrate the feasibility of quantitatively estimating drowsiness level using ICA-based multistream EEG spectra. The proposed ICA-based method applied to power spectrum of ICA components can successfully (1) remove most of EEG artifacts, (2) suggest an optimal montage to place EEG electrodes, and estimate the driver's drowsiness fluctuation indexed by the driving performance measure. Finally, we present a benchmark study in which the accuracy of ICA-component-based alertness estimates compares favorably to scalp-EEG based.

ANN based Joint Time and frequency analysis of EEG for detection of driver drowsiness

“…A significant decrease was observed in the levels of beta, alpha, theta activities within parietal, temporal and occipital lobe by other researchers [14][15][16][17][21][22] . However a quantitative approach to drowsiness state was attempted in this study.…”

“…A group of researchers have used independent component analysis (ICA) and Log Power Spectrum of EEG by using Fast Fourier Transform (FFT) to classify the driving performance into active and drowsy [16][17] . It has been reported by a group of researchers that there is a significant (ρ < 0.05) change in the Beta, alpha and theta activities within the temporal lobe.…”

“…A neuro-fuzzy information system is proposed to detect drowsiness level based on EEG signal. Although EEG analysis was extrapolated to various applications of drowsiness detection erstwhile, the quantification of drowsiness states is still not yet a reality [16][17] . This study bridges this gap and proposes a new index to quantify the drowsiness using the features of EEG signals.…”

In-Vehicle Stress Monitoring Based on EEG Signal

“…Artificial intelligence and Statistical methods have also been applied in a simulator---based study in 2005, [36] here the drivers have exposed to a drifting---task in a simulator. Here, the performances have measured by a moving-average of the lane position deviation.…”

Drowsiness detection for single channel EEG by DWT best m-term approximation

“…Previous studies commonly used Fast Fourier Transform (FFT) or Discrete Wavelet Transform (DWT) (Akay, 1998;Correa and Leber, 2010;De Carli et al, 1999;Liang et al, 2006;Subasi, 2005) as a feature selection technique, while Mahalanobis distance (Lin et al, 2010), independent component analysis (Lin et al, 2005) or neural network were used as classifiers (Subasi, 2005). Correa and Leber (2010) also used an artificial neural network to classify drowsiness through the extracted characteristics from an EEG signal using wavelets and Fourier spectrum.…”