ABSTRACT

Drowsiness is the state where a person is almost asleep or very lightly asleep. It refers to an
inability to keep awake. In this thesis drowsiness and sleepiness are considered synonymous, but
the term drowsiness will be used. Another concept commonly used is fatigue, which is an
extreme tiredness that comes from physical or mental activity. Drowsiness can also be described
by the grade of wakefulness or vigilance. Wakefulness is the same as alertness or a state of sleep
inability, whereas vigilance can be defined as watchfulness or a state where one is prepared for
something to happen. There are several factors which affect the grade of wakefulness. The time
spent to carry out a task (time on task) and the amount of sleep during night are the most
important factors. Other factors which are responsible are the amount of light, sound,
temperature and oxygen content. Motivation and monotony of the task will also have an effect on
the grade of wakefulness. Drowsiness can be measured by using physiological measures,
performance measures and self-reportor expert ratings.
In the present trend suggests driving and navigation support systems are getting
importance because it is crucial in supporting drivers in several conditions in automobile
industry. It is important for driving support systems to detect the status/activity of drivers
consciousness. Detecting onset of driver fatigue could prevent the accidents caused by drowsy
driving. It is proposed to detect/analyze the driver fatigue by the application of dedicated
physiological indicators such as electroencephalography and also facial features such as eyelid
movements based on template matching using neural network technique for closed eyes versus
opened eyes. EEG signal is one of the most predictive and reliable measurements by the analysis
of alpha, beta and theta band power etc., which are considered as direct factors associated with
human drowsiness. The parameters such as blink duration and opening time changes reliably
with increasing drowsiness in this system Biomedical Workbench in LABVIEW Biomedical
Toolkit used for capturing analyzing the data for this application.
This work, presented an analysis of time, spectral features of brain rhythms to identify
difference between awake and drowsy state in EEG signals. Then it was proposed an
automatic method to distinguish Awake State from Drowsy State using EEG features. The
advantage of using only one EEG channel out of 16 channels is that the driver would use

only one pair electrodes in the head instead of using an uncomfortable cap with
numerous electrodes. The EEG signal reflects the loss of alertness. well, so it is
appropriate to detect drowsiness and prevent accident.

ii

ACKNOWLEDGEMENT
I express my sincere gratitude to the Punjab Technical University, Jalandhar for giving me the
opportunity to work on the thesis during my final year of M.Tech. Thesis work is an important
aspect in tha field of engineering. I would like to thank Dr. R.K. DHAWAN,
PRINCIPAL,R.I.E.T, PHAGWARA for their kind support.
I also own my sincerest gratitude towards Er. Sukhpreet singh Bains supervisor for his valuable
advice and healthy criticism thoughtout my thesis which helped me immensely to complete my
work successfully.
I would like to thank my co-guide Mr. Karamjeet Singh and Dean Academics and member of the
Departmental Reserch Committee for their valuable suggestion and healthy criticism during my
presentation of work.
The constant guidance and encouragement received from Dr. Naveen Dhillon, Head of
Department of Electronics and Communication Engineering,(RIET), has been of great help in
carrying our present work and is ackonowledged with reverential thanks.
I would also like to thank everyone who has knowingly helped me throughout my thesis.
Last but not the least, a word of thanks for the authors of all those books and papers which I have
consulted during my thesis work as wall as for preparing the report.
At the end thanks to all almighty who made me capable of undertaking and completing this great
job.

SIMRAN JEET SINGH

1269612
Ramgarhia Institute of Engineering and Technology
phagwara

iii

LIST OF FIGURES
Figure No.

Caption

Page No.

1.1

change in EOG Potential when

1.2
1.3

looking 30 degree to the right

Electrod Placement
Defination of blink duration,T,

1.4

in EOG
EEG waves

1.5

Electrod placement

1.6

EEG

1.7

condition
EEG Pattern in somnolent

2.1

condition
Somnolence detection method

4.1

Placement of electrods on

4.2

scalp
Take samples utility

4.3

IMFS extracted using EMD

4.4

process
Showing the signal foe 16

Pattern

in

awake

channels for only 22 inputs

4.5

(subject 1)
Showing EEG

4.6 (a)

subject 2
EEG signal at time 22 minute

4.6(b)

11 seconds
EEG signal at time 20 minute

4.6(c)

51 seconds
EEG signal at time 18 minute

4.6(d)

31 seconds
EEG signal at time 16 minute

4.6(e)

22 seconds
EEG signal at time 12 minute

signal

iv

for

4.7

31 seconds
Variance-Covariance

EEG

signal

LIST OF TABLES
Figure No.
1.1
4.1

Caption
Page No.
Modified version of KSS
Showing the data for 16 channel

4.2
4.3

EEG only 25 samples

Showing the signal for subject 2
Parameters calculated EEG

4.4

signal at 22 minute 11 seconds

Parameters calculated EEG

4.5(a)

signal at 20 minute 51 seconds

Parameters calculated EEG

4.5(b)

signal at 18 minute 32 seconds

Parameters calculated EEG

4.5(c)

signal at 16 minute 22 seconds

Parameters calculated EEG

4.6

signal at 22 minute 11 seconds

Calctlated varience-Covariance
matrix

LIST OF ABBREVIATION
Abbreviation

Deseription

EEG

Electroencephalogram

NASTSA

National Highway Traffic Safety Administration

SEM

Slow sleepiness scale

EOG

Electrooculogram

SSS

Stanford sleepiness scale

KSS

Karolinska sleepiness scale

vi

EMD

Empirical Mode Decomposition

BCI

Brain Computer Interface

BMI

Brain Machine Interface

IMF

Intrinsic Mode Functions

STFT

Short Time Forier Transform

MCT

Means Comparison Test

VCT

Variances Comparison Test

MLP

Multi-layer Perceptron

EIV

Error-in Variables

LPC

Linear Predictive Coding

MFCC

Mel-Frequency-Cepstral-Coefficients

GMM

Gaussion Mixture Model

ANN

Artficial Neural Network

MGF

Magnified Gradient Function

ICA

Independent Component Analysis

SBP

Standard Back Propagation

LVQ

Learning Vector Quantization

SDLP

Standard Deviation Of Lateral Position

TCC

Time To Line Crossing

HPF

High Pass Filter

LPF

Low Pass Filter

vii

ORGANIZATION OF THESIS
CHAPTER 1 :- The first chapter provides the introduction to the concept, element, its
application, working and structure of proposed system.
CHAPTER 2 :- The second chapter briefly reviews the literature and previous work done.
CHAPTER 3 :- The third chapter gives the problem definition
CHAPTER 4 :- The forth chapter gives proposed algorithm
CHAPTER 5 :- The fifth chapter discusses the implementation of various tools of application
software for simulation and testing part of thesis
Finally, concluding thesis in sixth chapter with future scope.

viii