跌倒偵測是醫療照護領域上一道重要議題，跌倒後通常會在心裡以及生理上造成影響，特別是對於老人家的影響情況更是嚴重。在本研究中，我們提出了利用智慧型手錶來實現一跌倒偵測系統，所設想的情境為監護老人家於住家是否有跌倒情事發生，當確立跌倒事件發生後，我們設計一套有三個救援步驟的流程：第一步為以聲音訊息提示使用者是否需要救助，如果不需要，使用者即手動取消救助機制；第二步為確認需要救援後，會發送GCM通知給照護人員，讓照護人員第一時間得到訊息；第三步為照護人員收到通知後，可以通過重建事件發生功能，得知推測傷患跌倒情形及位置，讓傷患盡速得到醫療照護。為了實現以上程序，首先我們設計一套基於使用智慧型手錶的跌倒偵測演算法，所設計之演算法使用加速度數值來判定與分析跌倒進程，並利用SMV概念來判斷使用者是否跌倒以及是否需要救助。本系統亦利用iBeacon室內定位技術及行人航位推算概念來回報使用者在室內的路徑軌跡資訊，該資訊可供照護或照護人員能推估跌倒情事可能發生之原因，我們的實驗成果應證所設計之方法能有效地偵測跌倒偵測及重現事故。

Fall detection is an important issue in the field of medical care, which usually affects her heart and physiology, especially for the elderly. In this study, we proposed the use of intelligent watches to achieve a fall detection system, the scenario envisaged for the elderly in the home whether there is a fall occurred, when the establishment of the fall after the incident, we have designed a set of three rescue steps The first step is to prompt the user if the need for assistance, if not, the user manually cancel the rescue mechanism; the second step to confirm the need for rescue, will send a GCM notice to the care workers, so that the care of the staff The third step for the caregiver received a notice, you can rebuild the event function, that the risk of falls and the location of the situation, so that the injured as soon as possible medical care. In order to achieve the above procedures, we first design a set based on the use of smart watch fall detection algorithm, the algorithm used to calculate the acceleration value to determine and analyze the fall process, and the use of SMV concept to determine whether the user falls and need Rescue. The system also uses the iBeacon indoor positioning technology and the pedestrian deadline concept to report the user's trajectory information in the room. The information is available for care or caregivers to estimate the possible causes of falls. Our test results should be The design method can effectively detect the fall detection and reproduce the accident.

目錄	V
圖目錄	VI
表目錄	VII
第一章 緒論	1
第二章 相關研究	5
2.1固定式跌倒偵測系統	5
2.2攜帶式跌倒偵測系統	8
第三章 系統設定	11
3.1老人端服務	12
3.2伺服器端服務	14
3.3照護人員端服務	15
第四章 跌倒偵測模組	17
4.1跌倒偵測階段	17
4.2判定跌倒方向階段	21
第五章 實作成果與實驗	24
5.1實作之使用情境	24
5.2 iBeacon室內多點定位	25
5.3 預處理模組	27
5.4 軌跡產生	32
5.5 跌倒偵測實驗	33
第六章 結論	42
參考文獻	43
附錄一 英文論文 45

圖目錄
圖1: 系統架構圖	11
圖2:從穿戴式手錶取的一跌倒例子加速度圖	18
圖3 跌倒偵測流程圖	21
圖4:傷患跌倒在Impact跌倒之位移狀況	23
圖5:照護人員端手機顯示近三分鐘路徑	25
圖6:單顆iBeacon的狀況模擬圖及三顆iBeacon的狀況模擬圖	26
圖7: 感測裝置水平擺放	31
圖8: 感測裝置水平擺放	31
圖9 : 三種指定圖型路徑測試結果	33
圖10: 正面跌倒加速度At_impact_limit切割結果圖	36
圖11: 正面跌倒加速度Az_impact_limit切割結果圖	37

表目錄
表1:硬體規格表	34
表2: At_impact_limit設置與起始點找尋之結果圖	37
表3:Az_impact_limit設置與起始點找尋之結果圖	37
表4: 效能參數分類表	38
表5:三名受試者部分實驗結果表	40
表6:三個受試者之效能評比	41
表7:不規則方向跌倒結果	41

[1]	中華民國人口推估（105至150年），檢自:國家發展委員會http://www.ndc.gov.tw/Content_List.aspx?n=84223C65B6F94D72

[2]	徐慧娟、吳淑瓊、江東亮，「跌倒對社區老人健康生活品質的影響」，中華衛誌，Vol.15，No.6，1996。
[3]	林茂榮(Mau-Roung Lin) 、 王夷暐(Yi-Wei Wang)，社區老人跌倒的危險因子與預防，臺灣公共衛生雜誌
http://www.airitilibrary.com/Publication/alDetailedMesh?docid=10232141-200408-23-4-259-271-a

[4]	H. Nait-Charif and S. McKenna, “Activity summarization and fall detection in a supportive home environment,” in Proceeding of 17th ICPR, 2004, vol. 4, pp. 323–326.
[5]	Hui-min Qian, Yao-bin Mao, and Zhi-quan Wang, “SVM-based abnormal activity detection for home care,” in Proceeding of IEEE 7th World Congress on Intelligent Control and Automation, Chongqing, June 2008, pp. 3766-3771.
[6]	S.-G. Miaou, Pei-Hsu Sung, and Chia-Yuan Huang, “A Customized Human Fall Detection System Using Omni-Camera Images and Personal Information,” in Proceeding of 1st Transdisciplinary Conference on Distributed Diagnosis and Home Healthcare, Arlington, April 2006, pp. 39-42.
[7]	Majd Alwan, Prabhu Jude Rajendran, Steve Kell David Mack, Siddharth Dalal, and Robin Felder, “A Smart and Passive Floor-Vibration Based Fall Detector for Elderly,” in Proceeding of 2nd IEEE International Conference on Information and Communication Technologies, Damascus, 2006, pp. 1003-1007.
[8]	Andrew Sixsmith, N. Johnson,”A smart sensor to detect the falls of the elderly” EEE Pervasive Computing ( Volume: 3, Issue: 2, April-June 2004 )
[9]	Oscar Almeida，Ming Zhang，Jyh-Charn Liu，” Dynamic Fall Detection and Pace Measurement in Walking Sticks”，manuscript 2017
[10]	Solves a formulation of n-D space trilateration problem using a nonlinear least squares optimizer，https://github.com/lemmingapex/trilateration

[11]	Paola Pierleoni,Alberto Belli,Lorenzo Palma,A High Reliability Wearable Device for Elderly Fall Detection
[12]	Meng-Shiuan Pan and Hsueh-Wei Lin，” A Step Counting Algorithm for Smartphone Users: Design and Implementation”in Proceeding of IEEE SENSORS JOURNAL, VOL. 15, NO. 4, APRIL 2015
[13]	Meng-Shiuan Pan and Guan Ying Lee,”An Indoor Navigation System for Smartphone User”