Nuclearstar/Drowsiness_Detection

Drowsiness_Detection

This repository consists of the two implementations for driver drowsiness detection via

1. Eye monitoring being it closed or opened (Detection_for_images).
2. Real-time video stream (Detection_for_videos).

Code Requirements

1. pip install dlib
2. pip install skimage
3. pip install scipy
4. pip install numpy
5. pip install cv2
6. pip install imutils

Dataset

You can download a trained facial shape predictor from: here

Procedure in Implementation 1

Python code: Simple app to detect the status of eye being closed or opened (sleep_detection.py (test1.jpg and test2.jpg are for testing))

Run using the command, python sleep_detection.py /path/to/shape_predictor_68_face_landmarks.dat test1.jpg

Procedure in Implementation 2

To run the code, type python Drowsiness_Detection.py

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 lesst than 0.25, Alert is generated.

Relationship

Summing up

For more information look here.

References and Credits:

1. This implementation is inspired by Akshay Bahadur's project.
2. This implementation also took inspiration from Taha Emara's work.