Alertness Monitoring System for Vehicle Drivers using Physiological Signals

Alertness Monitoring System for Vehicle Drivers using Physiological Signals

Anwesha Sengupta (Indian Institute of Technology Kharagpur, India), Anjith George (Indian Institute of Technology Kharagpur, India), Anirban Dasgupta (Indian Institute of Technology Kharagpur, India), Aritra Chaudhuri (Indian Institute of Technology Kharagpur, India), Bibek Kabi (Indian Institute of Technology Kharagpur, India) and Aurobinda Routray (Indian Institute of Technology Kharagpur, India)
DOI: 10.4018/978-1-5225-0084-1.ch013
OnDemand PDF Download:
List Price: $37.50


The present chapter deals with the development of a robust real-time embedded system which can detect the level of drowsiness in automotive and locomotive drivers based on ocular images and speech signals of the driver. The system has been cross-validated using Electroencephalogram (EEG) as well as Psychomotor response tests. A ratio based on eyelid closure rates called PERcentage of eyelid CLOSure (PERCLOS) using Principal Component Analysis (PCA) and Support Vector Machine (SVM) is employed to determine the state of drowsiness. Besides, the voiced-to-unvoiced speech ratio has also been used. Source localization and synchronization of EEG signals have been employed for detection of various brain stages during various stages of fatigue and cross-validating the algorithms based in image and speech data. The synchronization has been represented in terms of a complex network and the parameters of the network have been used to trace the change in fatigue of sleep-deprived subjects. In addition, subjective feedback has also been obtained.
Chapter Preview


Driver fatigue is one of the major causes of road and rail accidents across the world that often result in death or serious injury. Crashes caused by tired drivers are most likely to happen when driving happens in long and monotonous roads, or when the driving schedule interferes with the circadian rhythm of the driver, or when the driver has had poor quality or quantity of sleep. Alcohol or certain medications have also been pointed out as probable causes.

Railway accidents are also a potential threat to the safety and lives of passengers. Besides, disregard of signal by locomotive drivers at level crossing gates can endanger the lives of road users. Increase in rail traffic leads, congestion of railway routes necessitating frequent interaction of train drivers with the traffic controllers, increase of locomotive speeds etc. has emphasized the need for train drivers to stay alert at all times. Reports indicate that of all the accidents that pose a threat to railway safety and management, many can be attributed to human errors, including performance of the driver.

Lack of adequate sleep can lead to depreciation in mood (irritability, anxiety, lack of motivation), health (increased blood pressure and increased risk of heart attack) and performance (lack of concentration, drop in attention/ vigilance, increased reaction time). Hence sleepiness impairs the ability to execute attention-based activities such as driving.

Alertness level in human beings can be assessed using different measures such as Electroencephalogram (EEG)(M. Bundele & Banerjee, 2009; Cajochen, Zeitzer, Czeisler, & Dijk, 2000), blood samples(Penetar, McCann, & Thorne, 1993), ocular features(Arief, Purwanto, Pramadihanto, Sato, & Minato, 2009), speech signals(Dhupati, Kar, Rajaguru, & Routray, 2010)and skin conductance(M. Bundele & Banerjee, 2009). Methods based on EEG signals and blood samples have been reported to be most accurate for estimating the state of drowsiness (Cajochen et al., 2000; Matousek & Petersen, 1983). However, these methods are contact-based and have restricted feasibility of implementation in practical scenarios. There have been some studies (Hanowski & Bowman, 2008; Inchingolo & Spanio, 1985; Smith, Shah, & da Vitoria Lobo, 2003; Wierwille, Wreggit, Kirn, Ellsworth, & Fairbanks, 1984; Zhu & Ji, 2004) with regard to eyelid movements such as blink frequency, Average Eye-Closure Speed (AECS), PERCLOS, eye saccade etc. as quantitative measures of the alertness level of an individual. Of them, PERCLOS is reported to be the best and most robust measure for drowsiness detection (Bowman, Schaudt, & Hanowski, 2012; Hammoud, Witt, Dufour, Wilhelm, & Newman, 2008) whereas saccadic movements are reported to be the best indicator of the alertness level (Ueno & Uchikawa, 2004). PERCLOS is based on eye closure rates whereas Saccadic Rate (SR) is based on fast eye movements.

Key Terms in this Chapter

Saccades: Saccadic movements are the ballistic movements of both eyes in the same direction which the subject is shifting the point of gaze. Eye saccadic movements are the one of the fastest movements human body can make.

Saccadic Ratio: The ratio of peak saccadic velocity to the saccadic duration.

Clustering Coefficient: Measure of segregation in brain networks. The fraction of triangles around an individual mode in the network; equivalent to the fraction of the node’s neighbors that are also neighbors of each other.

Fatigue: A physical and/or mental state of being tired and weak. May be caused by overwork, lack of sleep, anxiety, boredom, over/underactive thyroid glands, depression or certain medications.

PERCLOS: The percentage time where eyes are occluded at least by 80%.

Visibility Graph: A graph for a set of points in the Euclidean plane, where each node stands for the location of a point, and each edge represents a visible connection between them.

EEG: The electrical activity of the brain, recorded for a short period of time along the scalp. Measure of fluctuation of voltage resulting from ionic current flows within the neurons of the brain.

Entropy: Average amount of information contained in a sample drawn from a distribution or data stream. Measure of uncertainty of the source of information.

Characteristic Path Length: Measure of functional integration in brain networks. The average shortest path length between all pairs of nodes in the network.

Complete Chapter List

Search this Book: