Design and Implementation of a Real-time System for Continuous Monitoring the Heart-rate

[[abstract]]近年來，已開發國家均面臨高齡化現象與人口減少問題的挑戰，因而自主健康管理也逐漸受到重視。目前市面上許多監測生理訊號儀器皆以侵入式或是接觸式為主，設備龐大不易攜帶，並且需要專業的技術人員操作，價格也相對較高，故很多學者及本研究均透過非接觸式且較低成本的視訊量測方式，監測生理訊號。本研究提出一量測心率的流程技術，包括捕捉整臉視訊、濾除眼睛區塊、分離出綠光信號、及準位歸零調整，此流程技術能達到即時連續量測、使用整臉視訊可以增加量測距離及減少誤差。本研究實驗以webcam所量測出之心率，再與通過認證的SPO2機器做誤差驗證進行比較，經由統計18名測試者的有效樣本，證明採用本研究提出的流程技術，心率量測距離可以拉長到150公分、平均誤差值每分鐘在4個心跳之內。 In recent years, since developed countries have been facing the challenges of the aging phenomenon and the depopulation problem, using various instruments to measure human vital signs for health management purpose receives much attention. Many instruments for monitoring physiological signals using invasive or contact means can easily be found in the market. However, most of these instruments require professional technicians to operate and are usually expensive and difficult to carry. Therefore, many researches, including the presented study, try to adopt various webcam-based measuring methods with non-contact and low-cost features to monitor human vital signs. In this paper, we propose a heart rate inspecting technique to implement a real-time continuous monitoring system. The system includes the procedures of capturing full-face video, removing eye areas from the video, extracting green-light component from the video signal, and resetting signal level to zero. The proposed technique allows real-time continuous computation with longer measuring distance. In addition, it results in less average error than the conventional method using forehead area video. Lastly, we have measured the heart rate of 18 volunteers by using web camera and a proved SPO2 device simultaneously for comparison. The results show that heart rates can be measured from a distance of 150 cm and the average error value is within 4 heart beats per minute when the proposed technique is adopted