Heart Rate Measurement Using Video in Different User States for Online HCI Applications

In this paper we implemented an unobtrusive and non-invasive method to measure pulse rate and heart rate variability. A Ballistocardiography technique has been used which describes ballistic force applied by heart on blood vessels. Ballistocardiography depicts repetitive motion in human body against blood flow because of the ballistic force. In this process videos of human head have been recorded and feature points using Lucas - Kanade point tracker are calculated. After filtering the resultant signal to remove noise (unwanted signals) trajectories are decomposed into elementary components. Component that best corresponds to pulse frequency is selected. Detected peaks correspond to beats in the signal and hence the Heart Rate Variability is calculated. The algorithm was tested with many videos of 12 subjects (10 males and 2 females) with variations in skin color, sex, subject's state. We found that user state and activities affect Heart Rate Variability which can be calculated through our algorithm. It is concluded that extensive research and development of this unobtrusive and non-invasive technique can yield new tools for various online and adaptive HCI applications.