Project File




Start: Sep 03, 2018 - Ongoing

Image Processing, Machine Learning

Road accidents are suspected to be a primary concern nowadays and the key cause is the fatigue ....




Road accidents are suspected to be a primary concern nowadays and the key cause is the fatigue level of the driver.The system will track driver’s facial features like fast blinking, eye closures, head poses and yawning. We will try to achieve accuracy during the day, at night and under illuminated conditions through night vision camera.

Classifiers like Haar Cascade Classifiers or HOG + linear SVM will be chosen depending on their accuracy and shortest time of computation. EAR (eye aspect ratio) and facial expressions will be used to sense the tiredness of the car driver which will stimulate the system to ring alarms and send warnings consequently. The technical facets of the project will rely on OpenCV and python for facial recognition with Raspberry pi for hardware port connections.

Scope of Work
The project “Driver’s Fatigue Level Detection System (DFLD)” was designed and developed to easily facilitate, monitor and help detect when a driver may be exhibiting diminished. Our system keeps an eye on the driver’s behaviour to identify any signs that the driver is losing concentration to the point where safety is being compromised. Our rationale was to create a system in a way to avoid accidents risks which are caused due to long hours, shift work, lack of sleep or sleep apnoea that may exhaust the driver and cause him/her to sleep while driving. Our aim was to make our system efficient and less costly.
The expectation from our project was to reduce road accidents caused by fatigueness, , especially the ones at night when the drivers are very sleepy and have to go for long deliveries. We believed that it would help save people’s lives. With this system , drivers could get a buzzer alarm as a warning to wake up incase they’re about to fall asleep. As a conclusion we could say that we were able to implement all of the aspects which include head tilts, yawns and eye closures. Also we were able to integrate a night vision camera. We had also planned on introducing a water spray to sprinkle water if the system detects eye closures, however after taking feedback from FYP-1 , we planned not to do that because it would have caused an accident if the driver spent time rubbing his eyes which had water, especially the ones who had to clean their glasses if they wore spectacles.
Project Video

Syed Salman H

Syed Salman Haider

Developer/ Programmer

My basic task is to understand the algorithms and then test which algorithms are best for our project. I'll be using OpenCV and python programming lan..

Fatima Z

Fatima Zain

Quality Assurance Executive

Checked the bugs and fixes in the program and integrated effective sounds for head titles, etc.

Hiba H

Hiba Haider

Python developer and quality manager

Wrote code for detecting yawns and sending necessary alerts and alarms for the driver. Did testing and debugging of the code. Did development, researc..

Project is available for license.

Register your interest in this project.
Project is available for sale.

Register your interest in this project.
Project is looking for 0$ investment.

Register your interest in this project.