DriveAlert 🚗💤

Real-time Driver Drowsiness & Yawning Detection using MediaPipe, OpenCV, and Text-to-Speech.
DriveAlert monitors a driver’s eyes and mouth through a webcam feed to detect signs of sleepiness or fatigue and issues both on-screen and voice alerts.

---

📹 Demo

📸 Screenshots

---

🛠 Tech Stack

• Python: Main programming language.
• OpenCV: For video stream processing.
• MediaPipe: For robust facial landmark detection.
• NumPy: For efficient numerical computations.
• pyttsx3: For text-to-speech alerts.

---

📋 Features

• Real-time detection of drowsiness and yawning
• Works offline (no internet required)
• Voice alerts for better safety
• Easy to run on any webcam-enabled device

---

⚙️ How It Works

1. Capture — The webcam feed is read in real-time using OpenCV.
2. Detect Landmarks — MediaPipe detects facial landmarks (eyes, mouth).
3. Calculate EAR & MAR —
   • EAR < threshold → eyes closed for a certain duration → drowsiness alert.
   • MAR > threshold → mouth wide open for a certain duration → yawn alert.
4. Alert — Visual and voice alerts are triggered using pyttsx3.

---

🚀 Installation & Usage

Clone the repository

git clone https://github.com/saichethana28/DriveAlert.git
cd DriveAlert

Install dependencies

pip install opencv-python mediapipe numpy pyttsx3

Run the program

python main.py

Press 'q' to quit the program.

📜 License

This project is licensed under the MIT License – see the LICENSE file for details.

📬 Contact

Created by V. Sai Chethana 📧 Email: [email protected]