Towards a low-complex breathing monitoring system based on acoustic signals

Abstract

Monitoring the breathing is required in many applications of medical and health fields, but it can be used also in new game applications, for example. In this work, an automatic system for monitoring the breathing is presented. The system uses the acoustic signal recorded by a standard microphone placed in the area of the nostrils. The system is based on a low-complex signal parameterization performed on non-overlapped frames. From this parameterization, a reduced set of real parameters is obtained frame-to-frame. These parameters feed a classifier that performs a classification in three stages: inspiration, transition or retention and expiration providing a fine monitoring of the respiration process. As all of those algorithms are of low complexity and the auxiliary equipment required could only be a standard microphone from a conventional Bluetooth Headset, the system could be able to run in a smartphone device. © 2013 Springer-Verlag Berlin Heidelberg.