Non-contact Breathing Rate Detection Based on Time of Flight Sensor.

Abstract

There are a growing number of methods to detect a person's breathing rate, but most techniques still either require contact with body skin or are usually uncomfortable to wear, too expensive and unfriendly for daily monitoring. The massive adoption of smartphones in recent years has created many opportunities to improve daily health monitoring. In this work, we demonstrated that off-the-shelf ToF lens on smartphones can capture a person's breathing rate while still. In addition, we proposed a method for extracting breathing rate from ToF signal and compared it with actual breathing rate obtained from temperature sensor. We evaluated the breathing rate accuracy of 6 people at rest, with a mean absolute error of 0.009Hz when considering different mean breathing rate conditions. Moreover, the mean absolute error percentage is 3.56% and the root mean squared percentage error is 6.64%, which is smaller than other methods of non-contact breathing rate detection in recent works.