Drowsiness Detection using Facial Landmarks and Eye Blinking

This project utilizes facial landmarks and eye blinking to detect drowsiness in real-time using OpenCV, Dlib, and NumPy libraries. The project focuses on analyzing the eye blink patterns of an individual and determining their level of drowsiness based on predefined thresholds.

Table of Contents

• Introduction
• Dependencies
• Execution Steps
• Additional Information

Introduction

Drowsiness Detection is an essential aspect of ensuring driver safety, especially in long-duration journeys. This project aims to detect drowsiness by analyzing facial landmarks and monitoring eye blink patterns. By utilizing computer vision techniques, the project provides real-time feedback on the driver's drowsiness level, allowing them to take appropriate actions to prevent accidents.

Dependencies

To run this project, the following dependencies are required:

• OpenCV (cv2) - For basic image processing functions.
• NumPy (numpy) - For array-related functions.
• Dlib (dlib) - For deep learning-based modules and face landmark detection.
• Imutils (face_utils) - For basic operations of conversion.

Make sure to install these dependencies before running the project.

Execution Steps

To execute the Drowsiness Detection project, follow these steps:

1. Clone the project repository:

git clone https://github.com/PMorey22/Drowsiness-Detection.git

2. Install the project dependencies:
pip install -r requirements.txt

3. Download the pre-trained shape predictor model:
You need to download the pre-trained shape predictor model file "shape_predictor_68_face_landmarks.dat". You can download it from the following link:

shape_predictor_68_face_landmarks.dat

Once downloaded, extract the file and place it in the project directory.

4. Execute the project:
python drowsiness_detection.py

5. Drowsiness Detection in action:
The program will start the camera feed and analyze the eye blink patterns in real-time. The output window will display the video feed along with the detected faces, facial landmarks, and the drowsiness status.

6. Terminate the program:
To stop the execution, press the 'Esc' key.