Non-contact pulse wave extraction based on imaging and matrix processing

Extraction of pulse wave is the key for Chinese Pulse Diagnosis. Recently, the digitalized Chinese Pulse Diagnosis has become a hotpot for Chinese medicine. However, the technology is based on contact diagnosis and may cause discomfort and deviation. The imaging photoplethysmography (IPPG) could obtain physiological signals (heart rate, blood oxygen, et al.) using ambient light. It could be used for non-contact digitalized Chinese Pulse Diagnosis. However, there are serious baseline drift and noises owing to the changes of ambient light intensity and movement of human. Therefore, it is difficult to extract useful pulse wave signals for Chinese Pulse Diagnosis. The paper demonstrates a method to acquire pulse wave based on imaging and matrix processing. It only needs a laptop with a webcam and acquiring the images of people's face. By constructing the composite form of the Fourier transform matrix and the smooth prior matrix, the image data are processed and the pulse wave can be extracted with the high frequency noise and baseline drift being removed effectively.