Research On Video-based Heart Rate Detection Under Illumination Variation Interference

With the improvement of living quality,the needs of disease control and daily monitoring are continuously increasing.The heart rate,as one of the vital signs of human beings,is the most effective and most direct indicator for human health.It is important for the diagnosis of chronic diseases such as heart failure,atrial fibrillation,and sleep apnea.At present,the research focus in the field of biomedical engineering is how to easily and effectively detect the heart rate information of the human body.Traditional heart rate instrumentation requires direct contact with the body.This manner may cause irritation and discomfort.Recently,the noncontact heart rate measurement techniques have attracted wide attention.Among them,the imaging photoplethysmography(iPPG)achieves a noninvasive,convenient,comfortable,easy to operate heart rate measurements.The iPPG technology is susceptible to movement artifacts and ambient light changes.Therefore,to deal with the changes in ambient light,we proposed a novel framework to effectively evaluate heart rate from webcam videos captured during illumination changing conditions.The main tasks can be summarized as follows:First,we collect the video sequences,and determine facial region of interest(ROI)and background ROI through face feature points detection and tracking algorithm for each frame image.The framework takes the assumption that both facial ROI and background ROI have the similar illumination variations and the background ROI can be treated as the denoising reference by using PLS to extract the underlying common illumination variation sources existing in both ROIs.Then,a number of intrinsic mode functions are decomposed by applying MEMD to the illumination-variation-suppressed facial ROI.Finally,the IMF candidate with the largest maximal amplitude will be determined as the HR IMF.After that,peak detection algorithm is performed to estimate the HR.In conclusion,the experimental results demonstrated that our proposed method is a promising solution for practical non-contact webcam-based HR measurement applications.