Evaluation of Video Compression for Facial Image Photoplethysmography in Time Series Space

Yanjun Sun, Mitsuhiro Ogawa

Abstract: Cardiac pulse measurement from video images is a promising method, which non-invasively accesses physiological parameters of the living body. Using this method of PPG (photoplethysmography) imaging was attempted of instantaneous heartbeat interval measurement by directly observing the PPG waveform in uncompressed and compressed formats. The results show that a non-negligible noise component arises from compressed videos. Our findings indicate uncompressed files can clean the signal.