🚗 Driver Drowsiness Detection Embeded System

📖 Table of Contents

1. Project Overview
2. Technologies & Hardware
3. System Configuration
4. Expected Benefits
5. License

📌 Project Overview

The Driver Drowsiness Detection System is an embedded system designed to analyze the driver’s face in real-time and determine drowsiness. By implementing this system, we aim to prevent accidents caused by drowsy driving and create a safer driving environment.

Features

🛠 Technologies & Hardware

• Embedded Board: STM32F10x
• Camera Module: ESP32-CAM
• Communication Methods:
  • ESP32-CAM ↔ STM32F Board: USART1
  • STM32F Board ↔ Bluetooth Module: USART2
  • Bluetooth Module ↔ Android: Bluetooth Communication
• Timer Usage:
  • TIM2: PIR sensor input processing
  • TIM3: Vibration motor control (PWM output)
• Drowsiness Detection Model:
  • Dataset: NTHU Drowsy Driver Detection Dataset
  • Model Architecture: CNN
  • Framework: TensorFlow Lite (converted to TFLite for embedded implementation)

⚙ System Configuration

1. ESP32-CAM: Captures the driver’s face in real-time and processes the image using the TFLite model.
2. STM32F Board:
   • Manages communication between ESP32-CAM, Bluetooth module, and Android Smartphone.
   • Handles PIR sensor input for detecting driver presence.
   • Controls the vibration motor using PWM output based on received drowsiness alerts.
   • //추가..
3. Bluetooth Module:
   • Connects to a smartphone app to display alerts and warnings.
   • Can be integrated with external alert devices (e.g., speakers, vibration motors).
4. Android Smartphone:
   • Receives Bluetooth alerts from STM32F Board.
   • Automatically plays music when a drowsiness alert is received.
5. Vibration Motor: Provides a physical alert when drowsiness is detected.
6. PIR Sensor: Detects the presence of the driver.
7. Light Sensor: Adjusts drowsiness detection sensitivity dynamically under varying lighting conditions.

🚀 Expected Benefits

• Enhanced Driver Safety: Prevents accidents caused by drowsy driving.
• Integrated Alert System: Utilizes visual, auditory, and haptic feedback for effective warnings.
• Automatic Music Playback: Helps keep the driver awake by playing music upon drowsiness detection.
• Optimized Embedded System: Uses a lightweight model for real-time analysis.

🏗 Future Improvements

• Enhance accuracy by training the model with additional datasets.
• Improve smartphone app with additional UI features.
• Incorporate vehicle interior environment data (temperature, lighting, etc.).

📜 License

MIT License