Micro-sleep is one of the most common causes of fatal car accidents

New drivers' fatigue monitoring algorithm

A prototype device using an algorithm measuring eyelid closure will be tested by Peugeot Citroën in real driving conditions to assess its usability for drivers’ fatigue detection.

The algorithm, developed by an École Polytechnique Fédérale de Lausanne (EPFL) student Marina Zimmermann uses a single camera framing the driver’s eyes to monitor how much closed or opened the eyes are.

The method builds on a known system of fatigue indication known as PERCLOS (percentage of eye closure) but tries to calculate it in real time while taking into account the fact that the driver might turn his head, wear glasses, drive at night or through tunnels.

The innovative solution of the EPFL’s Signal Processing Laboratory uses a small infrared camera placed behind the wheel. While developing the algorithm, Zimmermann had to create an eye analysing module that would be able to disregard possible light effects as well as the drivers’ different eye morphologies. Afterwards, she established a 3D profiling of the eye and eyelids to distinguish whether the eye is open or closed. Eventually, the methodology was optimised to fit within the car’s dashboard.  

"The proposed algorithm is sufficiently robust and simple to run on a standard camera. It will be able to combine the degree of eyelid-opening information together with other data - like yawning or head tilt- provided by the already existing face tracking system," said Jean-Philippe Thiran, the projects’ supervisor.

The first test have rendered promising results. The car manufacturer Peugeot Citroën has recently partnered with the research team to implement the technology into prototype vehicles to conduct extensive testing in real driving conditions. 

Micro-sleeps are the most dangerous consequences of drivers’ fatigue. According to some estimates, they cause nearly a third of highway accidents. Several tiredness detection systems are currently in use, based mostly on the detection of the loss of vehicle control.

Recent articles

Info Message

Our sites use cookies to support some functionality, and to collect anonymous user data.

Learn more about IET cookies and how to control them