Go back to .. Jose Maria Perez-Macias / Research Projects

Breathing Evaluation Using Video and Contactless Signals

Jose Maria Perez-Macias (Research & Algorithm Development)

Exploratory Research in Computer Vision, Optical Flow & Contactless Physiological Monitoring

Breathing Evaluation Using Video and Contactless Signals

Overview

Exploring novel approaches to assess breathing patterns by integrating video-based computer vision analysis with contactless sensor signals.

Background

Traditional respiratory monitoring often relies on sensors attached directly to the body, which can be obtrusive and disrupt natural sleep. Video-based methods, combined with advanced signal processing from ambient sources, offer a completely non-invasive alternative.

Aim

To develop robust techniques utilizing video streams and supplemental sensors to accurately measure respiratory rate, tidal volume variations, and abnormal breathing patterns without direct skin contact.

Methods

Employing computer vision algorithms, optical flow tracking of chest-wall displacement, and machine learning to extract respiratory waveforms from video, integrating data from other non-invasive sensors for enhanced clinical reliability.

Results

Initial tests indicate that combining video analysis with additional sensor signals significantly enhances breathing pattern detection and movement artifact rejection, providing a viable contactless alternative for sleep and respiratory monitoring.

Resources & Contact

This exploratory research investigates multi-modal fusion between video analysis (remote photoplethysmography and optical flow) and contactless mattress sensors.