Dual-Band Infrared Video-Based Measurement Using Pulse Wave Maps to Analyze Heart Rate Variability

摘要

Remote photoplethysmography (rPPG) enables us to capture the vital signs such as pulse rate, respiratory rate, oxygen saturation, and even heart rate variability (HRV) without any contact devises. Although the papers of rPPG mainly focus on the use of standard RGB camera, it cannot used for the cases in night or under the dim light conditions. Therefore, in this paper, we propose a novel noncontact method for monitoring HRV without visible lighting. The proposed method uses dual-band infrared videos to ensure robustness to fluctuations in illumination. The hemoglobin component is extracted via a simple projection from the dual-band pixel values in logarithmic space. We demonstrate the accurate extraction of pulse wave signals using pulse wave maps. As the results, we indicated the effectiveness of HRV monitoring in the situation under the dim light condition.