Improved Driver Drowsiness Monitoring System using Real-time Eye Blinking Method

Improved Driver Drowsiness Monitoring System using Real-time Eye Blinking Method


Aree A. Mohammed1, Mohammed Q. Kheder1 and Azhee W. Muhammed3

1 Faculty of Science and Science Educations, Computer Department, University of Sulaimani, Kurdistan Region, Iraq. 

e-mail: aree.ali@univsul.edu.iq

e-mail: mohammed.kheder@univsul.edu.iq

3 School of Basic Education, Computer Department, University of Sulaimani, Kurdistan Region, Iraq. 

e-mail: azhee.muhamed@univsul.edu.iq





Article info


Original: 19 Mar  2015
Revised: 18 May  2015
Accepted: 31 May 2015
Published online: 
20 Dec. 2015 


Key Words:
drowsiness detection
eye blinking
eye tracking
detection accuracy
     


Abstract

Drivers with a diminished vigilance level suffer from a marked decline in their perception; recognition and vehicle control abilities and therefore pose a serious danger to their own lives and the lives of the other people. According to the National Highway Traffic Safety Administration (NHTSA), about 100,000 crashes are the direct result of driver drowsiness each year. This is the reason why more and more researches are made to build automatic detectors of this dangerous state. In this paper, an efficient drowsiness detection system based on eye state (close and open) is developed.The camera should be fixed in front of the drivers to capture real time frames. After that, the results from the camera (new frames) are subject to the some vision-based algorithm to detect the eyes. Finally, the eye blink detection is applied to determine the state of eyes. There are two states: open state and close state. Based on eye state the warming alarm or telephone calling should be done to the drivers for preventing undesired accident. The proposed system will be tested with different light conditions, namely, day and night vision to show the performance in term of efficiency and accuracy. The results explain that the proposed system which is based on Android platform has a high accuracy rate (%97.2) for drowsiness detection especially during the night vision.
.






Ċ
Kewan Omer,
Dec 22, 2015, 1:08 PM