A Real-Time Vision-Based Heart Rate Measurement Framework for Home Nursing Assistance

In this paper, we propose a practical framework to realize accurate, real-time and contactless heart rate measurement, which is essential for home nursing and medical assistance. This framework is composed of the Second Order Blind Identification to decompose the hidden remote photoplethysmogram signal, a series of signal analysis methods to correctly calculate heart rate and a lightweight but accurate peak detection method to draw heartbeat curve. The results are compared to a home pulse oximeter and proved to be highly consistent with it under various interaction scenes such as daily conversation. The heartbeat curve generated by the proposed framework also shows agreement with the oximeter. This framework is robust with great potential for domestic health assistance and emotional care in an ageless aging society.