Non-contact heart rate assessment based on spectral analysis of video images

Abstract

Heart rate is one of the main physiological indicators of the body, and the parameters of heart rate variability reflect various aspects of the functional and psychoemotional status. Automatic determination of a person's condition based on video sequence analysis is an important problem in various areas related to ensuring the safety of production, air and transport communications, prevention of crimes and terrorist threats, etc. Therefore, an important task is to develop methods and algorithms for analyzing video images that allow remote monitoring of heart rate parameters. Purpose – development and software implementation of a method for non-contact assessment of heart rate based on spectral analysis of a video image of a person's face recorded using a traditional video camera. A method for non-contact measurement of heart rate has been developed, including the stages of data collection and preprocessing, spectral analysis of video images and analysis of the information obtained to calculate the results. The difference between the values of the average heart rate recorded using contact sensors and using the developed software does not exceed 3-4 beats per minute. The proposed approach can be implemented as part of various information systems where it is required to control the functional and psycho-emotional status of a person, for example, systems for operator’s status monitoring.