Extracting human breathing rate from the fusion of multiple piezo-resistive membranes

E-textile sensors present a real potential since they can be integrated into unobtrusive systems such as garments. They allow the monitoring of physiological parameters without disturbing the operator. Although they are mainly used in the medical field for personal health care, these sensors are increasingly used in our daily activities. In sports, at work or at home, embedded sensors are used for on-line human monitoring and for improving the quality of life. The work presented here concerns the processing and evaluation of new sensors for monitoring respiratory activity. Six piezo-resistive membranes are integrated into the backrest of a seat similar to the one you would expect on a commercial aircraft, allowing for non-obtrusive monitoring. The objective is to extract valid and reliable features from the fusion of these sensors in order to provide a non-obtrusive alternative to chest-straps for monitoring respiratory activity. Results show the developed fusion algorithm performs better than a worn reference (chest strap) both in nominal and agitated contexts for monitoring breathing rate.