Real-time vision based respiration monitoring system

Abstract

Respiration rate is an important indicator of an individual's health. Existing respiration monitoring systems are usually contact based, and so the sensing device is attached to the subject's body. This creates discomfort to the patient and can interfere with normal breathing. This paper presents a vision based contactless method of monitoring respiration rate. The method involves image and signal processing techniques to extract chest and abdominal movements' information from a sequence of video images recorded using a single video camera. The system provided a real-time respiration monitoring information consisting of a plot of respiration signal and the value of respiration rate. The respiration data could also be stored to allow for off-line analysis to be performed. A caparison of the developed method against a number of contact-based respiration monitoring systems confirmed its effectiveness. The system is currently being further developed to make its operation more robust.