An efficient hybrid algorithm for non-contact physiological sign monitoring using plethysmography wave analysis

Nowadays, a variety of studies have shown that a person’s physiological indicators can be reliably extracted from a distance through an RGB camera. For this purpose, our work suggests an effective hybrid approach based on photoplethysmographic signal (PPG) analysis for the continuous monitoring of physiological signs, including heart rate and respiration rate. Our approach presents dependable methods for image and signal processing that combine moving average filtering and successive variable mode decomposition methods (w.r.t. MAF-SVMD). These two latter approaches will be applied to the PPG signal in order to overcome drawbacks, filter the disturbances in the signal, and decompose the PPG signal. Additionally, the suggested method, MAF-SVMD, has much lower computational complexity and is more robust to the initial values. Our findings were contrasted with those of traditional techniques, including EMD and VMD methods. Between the proposed approach and the VMD algorithm, the derived respiratory rates show an average delta error of 0 and a coefficient correlation of 0.99; however, between the suggested technique and the EMD method, the relative delta value is equal to 7.71 and a coefficient correlation of 0.69. While the retrieved heart rate from each of the three methods shown is generally convergent to the same value.