顔動画から心拍数を推定する：顔追跡と色信号で動きの影響を抑える方法

A Robust Remote Photoplethysmography Method

Remote photoplethysmography (rPPG) is a method for measuring a subjects heart rate remotely using a camera. Factors such as subject movement, ambient light level, makeup etc. complicate such measurements by distorting the observed pulse. Recent works on this topic have proposed a variety of approaches for accurately measuring heart rate in humans, however these methods were tested in ideal conditions, where the subject does not make significant movements and all measurements are taken at the same level of illumination. In more realistic conditions these methods suffer from decreased accuracy. The study proposes a more robust method that is less susceptible to distortions and has minimal hardware requirements. The proposed method uses a combination of mathematical transforms to calculate the subjects heart rate. It performs best when used with a camera that has been modified by removing its infrared filter, although using an unmodified camera is also possible. The method was tested on 26 videos taken from 19 volunteers of varying gender and age. The obtained results were compared to reference data and the average mean absolute error was found to be at 1.95 beats per minute, which is noticeably better than the results from previous works. The remote photoplethysmography method proposed in the present article is more resistant to distortions than methods from previous publications and thus allows one to remotely and accurately measure the subjects heart rate without imposing any significant limitations on the subjects behavior.