Skip to content

A Fancy Machine learning project to detect driver's fatigue.

Notifications You must be signed in to change notification settings

prosperityai/Driver-Fatigue-Detection

Repository files navigation

Driver-Fatigue-Detection

Motivation:

The risk, danger and sometimes tragic results of drowsy driving are alarming. The National Highway Traffic Safety 
Administration conservatively estimates that 100,000 police-reported crashes resulted in an estimated 1,550 deaths,
71,000 injuries, and $12.5 billion in monetary losses due to driver fatigue.
This application detects blinks and alerts of the driver's drowsiness. It handles variable lighting conditions and 
works pretty well in day as well as in the dark. It uses histogram equalization and gamma correction to eliminate 
the effect of lighting.

Create a virtual environment and activate it:

virtualenv -p python env
/env/Scripts/activate

Install the required packages to run the project:

python -m pip install -r requirements.txt

Run blinkDetect.py:

python blinkDetect.py

A blink is supposed to last for 300 to 400 milliseconds. So, if the eye remains closed for more than 800-900 ms, we can say that the person is either drowsy or sleeping. On the other hand, if the eye reopens after just 100ms, its considered an invalid blink discarded.

P.S: Press 'r' key to reset drowsiness alert and 'esc' to exit.