Data Fusion to Develop a Driver Drowsiness Detection System with Robustness to Signal Loss

Abstract

This study proposes a drowsiness detection approach based on the combination of several different detection methods, with robustness to the input signal loss. Hence, if one of the methods fails for any reason, the whole system continues to work properly. To choose correct combination of the available methods and to utilize the benefits of methods of different categories, an image processing-based technique as well as a method based on driver-vehicle interaction is used. In order to avoid driving distraction, any use of an intrusive method is prevented. A driving simulator is used to gather real data and then artificial neural networks are used in the structure of the designed system. Several tests were conducted on twelve volunteers while their sleeping situations during one day prior to the tests, were fully under control. Although the impact of the proposed system on the improvement of the detection accuracy is not remarkable, the results indicate the main advantages of the system are the reliability of the detections and robustness to the loss of the input signals. The high reliability of the drowsiness detection systems plays an important role to reduce drowsiness related road accidents and their associated costs.

The effect of torque feedback exerted to driver's hands on vehicle handling – a hardware-in-the-loop approach

“…Other applications of the proposed system include teleoperation of vehicles and driver behavior investigations such as works done in Friedrichs and Yang (2010) and Samiee, Azadi, Kazemi, Nahvi, and Eichberger (2014).…”

Assessment of Driver’s Drowsiness Based on Fractal Dimensional Analysis of Sitting and Back Pressure Measurements

“…Other studies have assessed, classified and predicted drowsiness states ( Ji et al, 2004 , 2006 ; Sayed et al, 2012 ; Singh and Banga, 2013 ; Kusuma and Sunitha, 2014 ; Samiee et al, 2014 ). However, no definite or effective methods emerged from this research for determining when a warning signal should be presented to the driver.…”

Prediction of Point in Time with High Crash Risk by Integration of Bayesian Estimation of Drowsiness, Tracking Error, and Subjective Drowsiness

“…It is possible to predict the psychological rating on drowsiness using behavioral and physiological measures [24,28,29]. However, currently it is impossible to predict the point in time with high crash risk (crash in simulated driving) before such a crash occurs.…”

“…Sayed and Eskandarian [23] succeeded in achieving an accuracy of about 90% in classifying drivers as being sleep deprived or non-sleep deprived using driver steering data. Samiee et al [24] evaluated the arousal level classified as being alert or drowsy using vehicle dynamic data such as vehicle longitudinal position and duration of eye closures, and showed that the proposed method could differentiate between the alert and the drowsy states with an accuracy of more than 87.78%. Eskandarian et al [25] and McDonald et al [26] indicated the effectiveness of such vehicle-based measures for assessing drowsiness.…”

“…The lower value of EEG-MPF shows that the arousal level is decreased. EEG-α/β can be calculated as the ration of the sum of α-band (8-12 Hz) power and θ-band (4-7 Hz) power to the power of β-band (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) Hz) power. The higher this value is, the lower the arousal level gets.…”