Real-Time Heart Rate Estimation Using a Standard Camera

Abstract

Health indicators are among the most important ways to assess health status and predict sudden health relapses. Non-contact health monitoring is critical for overcoming limitations of traditional vital sign measurement methods, such as infection risks and patient discomfort during continuous monitoring. This study will present a method for real-time heart rate measuring without contact with people by using a standard webcam and applying processing on the extracted signal from frames of video, applying a Butterworth filter, Chrominance-based method algorithm to obtain a signal similar to the heart rate signal, then calculate the heart rate using two methods, the first peaks detection and second method by Fast Fourier transform (FFT) and compare between them. These Experiments were conducted indoors with a lighting source in front of participants, and the number of participants was 15. The results obtained from the proposed method were compared with those of the Pulse Oximeter, which was medically approved, and the final results of median absolute error (MAE) were 1.28 for the Peak detection method and 2.47 for the FFT method. These results demonstrate the potential of webcam-based heart rate monitoring as a reliable, non-invasive alternative for health assessment.