International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 09 Issue: 05 | May 2022
p-ISSN: 2395-0072
www.irjet.net
Driver’s Drowsiness Detection by Analyzing Yawning and Eye Closure Rahul K1, Raj Suriyan G2, Rajesh S3, Udhayakumar G4 123 Student,
Department of Electronics and Communication Engineering & SRM Valliammai Engineering College Professor, Department of Electronics and Communication Engineering & SRM Valliammai Engineering College ---------------------------------------------------------------------***--------------------------------------------------------------------4Associate
Abstract - Driver’s drowsiness is the major cause of
neural networks is accurate but those algorithms took a long time and they do a lot of math calculations which will increase the time of execution of the program as it will cause a significant delay in creating an alert signal.
accidents. In this project, we are addressing this issue by creating a system that would alert the driver if he is drowsy or sleepy. At first, The face region is detected and tracked in the captured video sequence utilizing computer vision techniques in the first step. The eye and mouthparts were extracted and analyzed for drivers’ drowsiness. It is done by calculating the Eye aspect ratio(EAR) and Mouth aspect ratio(MAR). Both EAR and MAR has threshold value, The EAR value will decrease if the eyes were closed and the MAR value will increase if the mouth was opened for a yawn. When these values cross their threshold the buzzer starts to alert the driver.
3. METHODOLOGY 3.1. Flowchart
Key Words: Open-CV, Dlib, Shape-predictor-68-facelandmarks ,EAR, MAR
1. INTRODUCTION Driving drowsy is a major problem in today’s world. About 12% of the major motor vehicle accidents were caused by driving drowsy. In this paper, we address this issue by creating a system that would alert the driver if he/she was found to be sleepy. Our system analyses both the eye and mouth to detect drowsiness. As the drivers found midnight to be a great time to drive, as there will be no traffic. As it takes a toll on their sleep cycles which is significantly causing them to fall asleep while driving. So this system will alert them if they fell asleep while driving. Our system is independent of the subject so it can be employed in commercial systems. The eye closure can be detected by analyzing the eye aspect ratio and the yawing can be detected by analyzing the mouth aspect ratio (MAR). The EAR threshold is set to 0.2 and the MAR threshold is set to 30. We can detect the early signs of fatigue if there is any change in these values. If the value of EAR keeps decreasing it means that the driver is closing his eyes and if the MAR value increases it means the driver is yawing. We implement this logic to detect the driver’s fatigue in this project.
3.2. Video Sequence The input video sequence from the camera is analyzed in the first step. The input can be from a webcam or CSI camera in raspberry pi. In this step the video sequence got from the camera is processed in the open-cv environment. The video sequence is converted from color to black and white, as the black and white images can be processed at a higher speed. This video sequence is now will be processed in the open cv environment.
1. RELATED WORK In previous works[3], the computational needed to detect drowsiness are quite large, also they use only static images which is rather time-consuming. Eye detection through EEG signals is fast but they lack accuracy. Usage of large
© 2022, IRJET
|
Impact Factor value: 7.529
|
ISO 9001:2008 Certified Journal
|
Page 3101