Noncontact heart rate measurement using high sensitivity camera in low light environment

摘要

In this paper, we propose the method of remote estimationrnfor heart rate (HR) and heart rate variability spectrogram (HRVS)rnby analyzing hemoglobin concentration from facial RGB imagesrnin low light environment. The emotion monitoring has a greatrnpotential in areas such as market research, safety measure,rnmedical, and robot systems and so on. Various methods werernproposed for the remote emotion estimation. Especially, thernmethods to estimate emotion by analyzing physiological signalsrnfrom RGB video are expected to be used practically. However,rnthese studies could not be applied in dark places wherernmonitoring is necessary such as infant’s bedrooms, crime pronernroads and car drivers. Therefore, the proposed method uses thernhigh sensitivity RGB camera capable of capturing videos at lowrnilluminance. As the result, we could measure HR with over 99%rnaccuracy and estimate HRVS with high accuracy in low lightrnenvironments of 10 lux that is the same brightness as the twilightrnwhere this low light environment is required in monitoring infant’srnbedrooms, crime prone roads, car drivers.