Drowsiness Detection using OpenCV

Detects eye movements and alerts when the object is drowsy

Applications

This can be used by riders who tend to drive for a longer period of time that may lead to accidents

Code Requirements

The example code is in Python (version 2.7 or higher will work).

Clone the repository

$ git clone https://github.com/dhrubajyoti89/drowsiness-detection.git

Open the drowsiness-detection Directory

$ cd drowsiness-detection

Make a virtual environment

$ pip install virtualenv

$ virtualenv drowsiness_detection_venv

$ cd drowsiness_detection_venv

Activate the virtual environment for Mac & Linux

$ source Scripts/activate

Activate the virtual environment for Windows

$ Scripts/activate

Dependencies

1. import cv2 (comes under opencv-python library)
2. import imutils
3. import scipy
4. import dlib

If you want to run through Jupiter Notebook :

1. install ipykernel

prerequisites

dlib library has some prerequisites such as Cmake and some C++ libraries. These prerequisites are essential otherwise the programme may throw RUN TIME ERRORS.

For Windows : Refer Here

For Mac and Linux : Refer Here

Go through the instructions in the above link and complete them first or else the below installation may throw errors

Install the dependecies

$ cd ..

$ pip install -r requirements.txt

Be patient it may take some time.

Run The Programme

After activating the virtual environment run :

$ python Drowsiness_Detection.py

Input

We can run the programme for live detection or even for a given video file (.avi file).

after running the programme

press 0 for live detection.

Or

press 1 to input a video file
Enter the path :

example path:(D:\Videos\testfile.avi)

The input video file must be of .avi format

Description

A computer vision system that can automatically detect driver drowsiness in a real-time video stream and then play an alarm if the driver appears to be drowsy.

Algorithm

Each eye is represented by 6 (x, y)-coordinates, starting at the left-corner of the eye (as if you were looking at the person), and then working clockwise around the eye:.

Condition

It checks 20 consecutive frames and if the Eye Aspect ratio is less than 0.25, Alert is generated.

Relationship

Summing up

For more information, Click Here

Contributions are welcomed !

First create an issue if you want to add new feauture or want to improve existing code. After that create a Pull Request.

Raise the PR to the TEST branch !

Make a folder named after Test , if the code is platform specific then please create subfolder named to that Operating system within that language folder. For example Test/Linux. Add README.md in it telling about compilation, installing dependencies and some screenshots of the updates.

Join the Slack channel for discussions 👋 : JOIN NOW