Non-Contact Vital Signs Monitor

Doppler Radar + Camera rPPG | Dual-Modality Contactless Health Monitoring

SGSITS Indore | Department of Biomedical Engineering | Minor Project 2025–26

---

What This Does

A low-cost dual-modality system that monitors two vital signs without touching the patient:

Vital Sign	Method	Sensor
Breathing Rate	Doppler Effect (chest wall motion)	HB100 10.525 GHz Radar + Arduino
Heart Rate	Remote Photoplethysmography (rPPG)	Standard Webcam + Python

Both are displayed simultaneously on a real-time Python dashboard. Breathing rate also appears on an embedded OLED display for standalone use.

---

Results

Metric	Our System	Commercial (TI AWR1642)
Heart Rate MAE	±3.2 BPM (5 trials vs. smartwatch)	±3–5 BPM
Breathing Detection	Normal / Tachypnea / Bradypnea / Apnea	Normal / Tachypnea / Bradypnea
Apnea Detection	Yes (8-second threshold)	Yes
Total Cost	Rs. 1,150	Rs. 30,000+

---

Hardware

Component	Specification	Cost
Arduino Uno R3	ATmega328P, 16 MHz	Rs. 500
HB100 Doppler Radar	10.525 GHz CW	Rs. 350
SSD1306 OLED	0.96", 128×64, I2C	Rs. 150
Breadboard + Wires	—	Rs. 150
Webcam	Built-in laptop / 720p USB	Rs. 0
Total		Rs. 1,150

---

Wiring

HB100 VCC  →  Arduino 5V
HB100 GND  →  Arduino GND
HB100 IF   →  Arduino A0

OLED VCC   →  Arduino 5V
OLED GND   →  Arduino GND
OLED SDA   →  Arduino A4
OLED SCL   →  Arduino A5

Arduino    →  Laptop USB (115200 baud serial)

---

Signal Processing

Radar Path (Arduino firmware)

HB100 IF Signal (5–50 mV, 0.1–0.5 Hz)
    ↓
10-bit ADC sampling
    ↓
50-sample moving average  ← low-pass, removes HF noise
    ↓
Peak detection (swing > 2 ADC units, 2s refractory period)
    ↓
Circular buffer (last 10 breaths)
    ↓
BR = 60000 / avg_interval_ms
    ↓
Apnea flag if no breath for > 8 seconds

rPPG Path (Python)

Webcam frame (720p, ~30 fps)
    ↓
Haar Cascade face detection (Viola-Jones)
    ↓
ROI extraction: forehead (top 15%) + left cheek
    ↓
Green channel mean (most sensitive to haemoglobin absorption)
    ↓
DC removal → Polynomial detrending (4th order)
    ↓
5th-order Butterworth bandpass filter (0.9–2.0 Hz = 54–120 BPM)
    ↓
Hanning-windowed FFT (20-second analysis window)
    ↓
Peak frequency × 60 → BPM
    ↓
Trimmed median stabilization (±1 BPM/step)

---

How to Run

1. Arduino (Breathing Rate Monitor)

Install required Arduino libraries (via Library Manager):

• Adafruit SSD1306
• Adafruit GFX Library

Open radar_breathing/radar_breathing.ino in Arduino IDE. Select Board: Arduino Uno | Port: your COM port. Click Upload.

2. Python (rPPG Heart Rate Monitor)

# Install dependencies
pip install -r requirements.txt

# Run
python rppg_heart_rate.py

Sit facing the webcam in a well-lit room. The system needs ~25 seconds to display a stable heart rate.

Press q to quit.

---

Repository Structure

contactless-vital-signs/
│
├── radar_breathing/
│   └── radar_breathing.ino     # Arduino firmware (breathing monitor)
│
├── rppg_heart_rate.py          # Python rPPG heart rate extraction
├── requirements.txt            # Python dependencies
└── README.md

---

Comparison with Commercial Systems

Parameter	This Project	TI AWR1642	Xethru X4
Cost	Rs. 1,150	Rs. 30,000+	Rs. 75,000+
Radar Frequency	10.525 GHz CW	77 GHz FMCW	7.29 GHz UWB
Breathing Accuracy	±2–3 br/min	±1 br/min	±1 br/min
Heart Rate	rPPG ±3.2 BPM	Radar ±3–5 BPM	Radar ±3 BPM
Range	~30 cm	1–3 m	0.5–5 m
Display	OLED + Dashboard	PC Software	PC Software
Power	USB 5V	12V adapter	5V USB

This system achieves the primary goal of affordable contactless monitoring at under 2% of commercial system cost.

---

Future Work

• Capacitive Contactless ECG — AD8232 with copper plate electrodes through clothing
• ML Cardiac Detection — 1D CNN + LSTM on radar BCG signal
• Wireless Transmission — HC-05 Bluetooth to mobile app (Flutter)
• Temperature — MLX90614 infrared sensor integration
• Sleep Apnea Screening — Overnight radar-based monitoring
• Clinical Validation — Bland-Altman analysis with larger sample size

---

Team

Name	Enrollment No.
Nirmal Verma	0801BM231041
Divyanshi Sharma	0801BM231023
Prashansa Sharma	0801BM231043

Guided by: Prof. Avni Jain & Prof. Gauri Gupta Department of Biomedical Engineering, SGSITS Indore

---

References

1. C. Li et al., "A Review on Recent Advances in Doppler Radar Sensors for Noncontact Healthcare Monitoring," IEEE Trans. Microwave Theory Tech., 2013.
2. W. Verkruysse et al., "Remote plethysmographic imaging using ambient light," Optics Express, 2008.
3. M. Z. Poh et al., "Advancements in Noncontact, Multiparameter Physiological Measurements Using a Webcam," IEEE Trans. Biomed. Eng., 2011.
4. G. De Haan and V. Jeanne, "Robust pulse rate from chrominance-based rPPG," IEEE Trans. Biomed. Eng., 2013.