# Adaptive Thermal Imaging Signal Analysis for Real-Time Non-Invasive Respiratory Rate Monitoring

## TL;DR

This paper introduces a real-time, non-invasive system for monitoring respiratory rate using thermal imaging and embedded hardware, achieving high accuracy in various conditions.

## Contribution

The novel contribution is an adaptive, embedded thermal imaging system that combines YOLO detection, Kalman filtering, and a MAD-hysteresis algorithm for robust respiratory rate estimation.

## Key findings

- The system achieved a mean absolute error of 0.57±0.36 BPM and root mean square error of 0.64±0.42 BPM across multiple experimental conditions.
- The proposed method outperformed peak-based and FFT spectral baselines in terms of error reduction across all tested scenarios.
- The system maintained accuracy under motion, thermal drift, and variations in distance and posture.

## Abstract

(1) Background: This study presents an adaptive, contactless, and privacy-preserving respiratory-rate monitoring system based on thermal imaging, designed for real-time operation on embedded edge hardware. The system continuously processes temperature data from a compact thermal camera without external computation, enabling practical deployment for home or clinical vital-sign monitoring. (2) Methods: Thermal frames are captured using a 256×192 TOPDON TC001 camera and processed entirely on an NVIDIA Jetson Orin Nano. A YOLO-based detector localizes the nostril region in every even frame (stride = 2) to reduce the computation load, while a Kalman filter predicts the ROI position on skipped frames to maintain spatial continuity and suppress motion jitter. From the stabilized ROI, a temperature-based breathing signal is extracted and analyzed through an adaptive median–MAD hysteresis algorithm that dynamically adjusts to signal amplitude and noise variations for breathing phase detection. Respiratory rate (RR) is computed from inter-breath intervals (IBI) validated within physiological constraints. (3) Results: Ten healthy subjects participated in six experimental conditions including resting, paced breathing, speech, off-axis yaw, posture (supine), and distance variations up to 2.0 m. Across these conditions, the system attained a MAE of 0.57±0.36 BPM and an RMSE of 0.64±0.42 BPM, demonstrating stable accuracy under motion and thermal drift. Compared with peak-based and FFT spectral baselines, the proposed method reduced errors by a large margin across all conditions. (4) Conclusions: The findings confirm that accurate and robust respiratory-rate estimation can be achieved using a low-resolution thermal sensor running entirely on an embedded edge device. The combination of YOLO-based nostril detector, Kalman ROI prediction, and adaptive MAD–hysteresis phase that self-adjusts to signal variability provides a compact, efficient, and privacy-preserving solution for non-invasive vital-sign monitoring in real-world environments.