TY  - JOUR
T1  - Minimization Collision and Robbery Framework for Your Vehicle
AU - Mohamed, M.A. AU - El-Den, Bassant M. 
JO  - International Journal of System Signal Control and Engineering Application
VL  - 9
IS  - 5
SP  - 107
EP  - 120
PY  - 2016
DA  - 2001/08/19
SN  - 1997-5422
DO  - ijssceapp.2016.107.120
UR  - https://makhillpublications.co/view-article.php?doi=ijssceapp.2016.107.120
KW  - Face recognition
KW  -eye tracking android
KW  -template matching
KW  -Arduino
KW  -servo motor
AB  - Driving vehicles is one of the amusing things that individually do, a side interest for another but in
the meantime is a hazardous instrument that could lead to death if used in the wrong way. Recently, the number
of vehicle’s theft had risen essentially which led to great losses at both individuals and establishment
(insurance agencies) levels. Thus, a great interest emerged as ways to protect vehicles from theft. This study
presents anti-robbery and protection framework that consists of two phases. In the first phase, face recognition
techniques are employed to identify the vehicle’s owner to secure the vehicle from burglary. Once recognized,
the bluetooth on the phone activated and signally connected to the arduino which is fitted inside the vehicle.
Then arduino connecting (shield, wires, servo motor) of the vehicle door lock and engine; opens the lock and
start the engine. Otherwise, if the user has not identified the door remains locked and the alert system is
activated. In the second phase, during driving the eye locomotion is observed using a mobile to reduce or limit
collision. In the case of proven snoozing, the reflexes of the driver are translated by the designed system to
warn the driver by a beep to save his life and also signal an alarm that stops the vehicle sequentially using its
brake system. The process is repeated every 6 sec to perceive if there is any distraction or somnolence that
might occur during driving or not. Three types of the database have been used to test the proposed framework
namely, face 94, ORL and Live database. The performance evaluations metric based on the False Rejection Rate
(FRR), False Acceptance Rate (FAR) and elapsed time have been used to assess the effectiveness of different
facial recognition techniques. Software results using MATLAB on a test set of photos have proved that the
Principle Component Analysis (PCA) technique has a superior performance. The executed framework uses the
android operating system in a smartphone to assist in detecting drivers under fatigue and alert driver under
sleepy conditions. Comparison different techniques, the Block Matching Algorithm (BMA) demonstrated a
superior performance for tracking driver’s eyes for limiting a collision in real-time according to elapsed time
parameter. Hardware implementation is executed using mitsubishi lancer vehicle and smartphone (Samsung
Galaxy S4) has been used for testing the proposed framework.
ER  -