透過非接觸的方式收集心跳訊號並運用於獨居老人的看護

Abstract

由於科技的進步，使得老年人口的比例上升，間接產生許多獨居老人沒人照顧的狀況發生。為了避免獨居老人在家中過世無人發現的案例再發生，我們透過本篇論文的方式來監控獨居老人的生理狀況。不僅可以立即發現獨居老人是否活者，更可以隨時監控獨居老人的心跳，如果出現異常的心跳可以發出警告請醫療單位前往照顧。獨立向量分析是一種分析訊號的方式，從混合的訊號中找出獨立的向量，可以做更進一步的分析，並且從中找出心跳的訊號。為了能夠辨認獨居老人是否活著，本論文利用人臉因為心跳而產生的顏色變化，以及人呼吸產生的振動來分辨此人是否活著。由於實驗的限制，利用照片來模擬過世的人來進行實驗。整合以上兩個功能，可以完成一個不侵犯獨居老人隱私，卻可以有效的照顧他們的方式。

Thanks to the technology become more advanced, the population of the elderly people become lager, which cause many elderly people live alone with others’ care. In order to avoid these people dying without anyone notice, this thesis provide a new method to achieve the elderly care. The method not only can found out whether the elderly are still alive immediately, but also measuring their heart rate. If his or her heart rates become unusual, someone can notice that immediately through this system. Independent component analysis, a method to analyze signals, can find the independent signals from the mixed signals, and further estimate the heart rate. To distinguish if the elderly is alive, this thesis use the variance on the face while a person’s heart is beating and a person’s move while he or she is breathing. We use picture of a human to simulate a dead person. Combine the method above we provide a method that didn’t invade the elderly’s privacy, but take good care of the elderly.