Drowsiness Monitor
Raspberry Pi · PPG, EAR, alarms
Embedded Systems · Edge Sensing
Real-Time Drowsiness Monitor
Two Raspberry Pi nodes combining PPG and eye-aspect-ratio analysis, TCP/IP communication, and GPIO/I²C alarms.
Source code and physical prototype
Open the files
01
Problem
The project acquires and analyzes PPG and camera-based EAR signals across two Raspberry Pi nodes and integrates the resulting alarm path.
02
Architecture
A PPG sensing node and an image/EAR analysis node communicate through TCP/IP and file IPC, with GPIO/I²C driving LCD and alarm outputs.
03
Hardware and Software
- Raspberry Pi
- C/C++
- OpenCV
- TCP/IP
- GPIO
- I²C
- LCD
04
Data Flow
PPG/camera sensing → per-node preprocessing → network/file IPC → fatigue-decision path → LCD/alarm output.
05
Methodology
- Integrated PPG-server and image/EAR-processing modules
- Documented two-node communication and state transfer
- Documented the LCD/alarm interfaces and prototype wiring
06
Results
- Two-node architecture implemented in the C/C++ source
- Data and alarm paths connected through TCP/IP, GPIO, and I²C
- LCD and sensor wiring shown in the original prototype photographs
Architecture and result views
What I worked on
- PPG server and image/EAR processing
- Two-node file IPC and network architecture
- Documentation of the LCD and alarm integration path
Code and results
- C/C++ source and architecture documentation
- Circuit, LCD, and prototype photographs
- TCP/IP, GPIO, and I²C data and alarm paths
Design scope and next steps
Available records: prototype, communication, and alarm paths. Follow-up measurements: latency, sensitivity, specificity, and false-alarm rate.
Project resources
A two-node Raspberry Pi prototype splits PPG and camera EAR processing, exchanges results over TCP/IP, and drives GPIO/I²C alarms.