2026-04-13T15:20:35Zhttps://recercat.cat/oai/request

oai:recercat.cat:10230/448112025-12-25T19:57:05Zcom_2072_6col_2072_452954

00925njm 22002777a 4500 dc Garcia Gómez, Guillem author 2020-05-26T08:28:52Z 2020-05-26T08:28:52Z 2018 Treball de fi de grau en Biomèdica Tutor: Federico Mateo Sukno, Alex Pereda The remote measurement of physiology is an interesting topic that has gained relevance in recent years as it can be a promising alternative to traditional methods to detect body signals. In particular, remote photoplethysmography, which consists in the distant measurement of the pulse from videos of subjects, has shown a lot of potential, successfully being used to measure the heart rate variability. Heart rate variability (HRV) is mostly regulated by the autonomic nervous system (ANS). The parasympathetic branch of the ANS, influences heart function when the subject is in a relaxed situation, increasing its variability and lowering the heart rate. On the other hand, the sympathetic branch is responsible for raising the heart rate and lowering the variability when facing a dangerous or stressful situation. In this project, we will explore the potential of remote photoplethysmography as a tool for non-contact measurement of stress. 20 subjects took part in an experiment designed to induce sates of relaxation and mental stress while their faces were being recorded with a camera. The pulse signals were extracted from the video and used to obtain a set of measures of HRV, then we compared the measures corresponding to rest and stress states. Our results indicate that the developed system could estimate the heart rate successfully, showing low error in comparison with the ground truth. We also found significant differences between some of the HRV measures for the relaxation and stress conditions. However, some of the measures obtained showed low correlation with the ground truth suggesting that the related HRV differences between rest and stress states may be originated by other factors. Nonetheless, this technique shows a lot of potential and could be used for real life applications. Heart rate variability Remote photoplethysmography Stress assessment Non-contact measurement Heart rate variability measuring from facial videos to detect stress