本論文設計一個人體生理健康預警分析系統，利用攝影機結合影像辨識追蹤技術，進行動態環境個人的定位。並透過IoT 穿戴式人體資訊感測器，即時將人體的生理資訊傳到藍牙5.0 的邊緣平台系統。並透過ECG 的波形PQRST 計算QT_duration(Q點到T 點之時間週期)，條件一為運動中的QT_duration(Q 點到T 點之時間週期)大於開始運動前的QT_duration (Q 點到T 點之時間週期)與條件二運動中QTc(正規化後的Q 點到T 點之波形)之斜率大於0.3，值0.3 是由本實驗26 個受測者中從臨床表現觀察得到的結果，同時符合這二個條件，系統會警示，可預警在運動中血壓的過度變化，以防猝死的發生。本論文主要有三個部分: (1)人體生理健康之分析(2)攝影機自動追蹤動態環境的個人定位(3)藍牙5.0 邊緣平台實現即時監控之預警系統。在人體生理健康分析方面，分析危害人體健康的因素有哪些，並利用IoT 感測器，將人的生理資訊收集並分析。在攝影機自動追蹤方面，透過演算法YOLOV3+deepsort 在動態環境下，辨識出人所在的環境區域並追蹤此人，直到出了這台攝影機的範圍，再由另一台攝影機繼續辨識與追蹤。在藍牙5.0 邊緣平台的設計方面，此系統可在無外接電源的環境下工作，讓系統可更方便佈建與攜帶。在傳輸距離上設計出一個有效的長距離接收天線，透過特殊的傳輸方式，在實際測試的結果，可讓每隻天線的傳輸距離達到800 公尺之遠。相較一般的藍牙5.0 最遠距離只有300 公尺，足足超過了2.5 倍的距離。再借由雲端的系統整合同步，即時監控人體的生理資訊，達到個人化即時預警的目的。

This paper designs a physical health early warning analysis system. Using cameras and image recognition tracking technology to locate individuals in dynamic environments. And through IoT wearable body information sensor, real-time transmission of physiological information and dynamic environmental information to the Bluetooth 5.0 edge platform system. And calculate the QT_duration (Q point to T point time period) from the waveform PQRST of ECG, Condition 1 is that the QT_duration (Q point to T point time period) of during exercise is greater than the QT_duration (Q point to T point time period) before starting the exercise. Condition 2 the slope of the QTc (normalized Q point to T point time period of waveform) during the exercise is greater than 0.3, and the value 0.3 is the result obtained from the clinical performance observations of 26 subjects in this experiment. While meeting these two conditions, the system will warn. Predict excessive changes in blood pressure during exercise to prevent sudden death.This paper has three main parts: (1) Analysis of physical health (2) Personal positioning of the camera to automatically track the dynamic environment (3) An early warning system for real-time monitoring of the Bluetooth 5.0 edge platform.In terms of physical health analysis, analyze what are the factors that endanger human health, and use IoT sensors to collect and analyze body physiological information.In terms of automatic camera tracking, through the algorithm YOLOV3 + deepsort in a dynamic environment, a person's environment area is identified and the person is tracked until it is out of the range of this camera, and then another camera continues to identify and track.With regard to the design of the Bluetooth 5.0 edge platform, This system to work in an environment without external power, which is convenient to carry and deploy. Design an effective long-range receiving antenna. Through a special transmission method, the actual test results can make receiving distance of each antenna reach 800 meters. Compared to the average Bluetooth 5.0, the maximum distance is only 300m, which more than 2.5 times. Then through the integration and synchronization of the cloud system, the physiological information of the body can be monitored in real time to achieve the purpose of personalized real-time warning.

目錄
中文摘要I 
英文摘要II 
目錄IV 
圖目錄VII 
表目錄X 
第一章緒論1 
1.1 研究動機1 
1.2 影響人體生理健康之因素3 
1.2.1 壓力造成的影響3 
1.3 IoT感測器ECG簡介4 
1.3.1 ECG量測設備簡介4 
1.3.2 ECG波形介紹6 
1.3.3 先前研究技術比較8 
1.3.4 HRV技術簡介9 
1.3.5 HRV-ECG介紹10 
第二章穿載裝置之系統探討設計13 
2.1 ECG穿載裝置之設計13 
2.2 人體體溫穿載裝置之設計14 
2.3 穿戴式裝置之FDA認證設計17 
2.3.1 硬體設計遇到之問題與解決17 
第三章個人環境影像追蹤與生理資訊整合系統之設計18 
3.1 影像辨識簡介18 
3.2 影像辨識演算法19 
3.3 影像追蹤演算法20 
3.4 個人動態環境影像追蹤定位流程圖22 
3.5 個人動態環境影像追蹤定位結果23 
3.6 個人動態環境影像追蹤與生理資訊整合實驗結果24 
第四章中繼邊緣平台系統架構26 
4.1 中繼邊緣平台系統之通訊方法說明26 
4.2 中繼邊緣平台系統之整體架構說明27 
4.2.1 中繼邊緣平台系統之天線設計28 
第五章中繼邊緣平台系統應用範例實測30 
5.1 中繼邊緣平台系統實測環境方法之簡介30 
5.2 中繼邊緣平台系統實測生理資訊之結果34 
第六章運動動態血壓預警與環境生理健康之分析研究38 
6.1 運動動態血壓之實驗流程38 
6.2 運動動態血壓預警之分析流程39 
6.3 運動動態血壓實測受測者生理基本資訊與臨床表現41 
6.4 環境生理健康之壓力實測結果分析47 
6.5 環境生理健康之各種環境下實測結果分析49 
第七章總結與貢獻 58 
第八章未來研究60 
參考文獻61

 
圖目錄
圖1.1論文研究主軸示意圖2 
圖1.2本論文使用之ECG模組實體圖5 
圖1.3本論文使用之ECG模組系統架構圖5 
圖1.4本論文使用之ECG實測波形圖6 
圖1.5心電圖PQRST示意圖7 
圖1.6 HRV量測波形之RRI參數10 
圖1.7 HRV量測之時域分析圖10 
圖1.8 HRV量測之頻域分析圖10 
圖2.1ECG moduleRecord Flow Chart 14 
圖2.2TD+工作原理示意圖15 
圖2.3TD+模組實體圖15 
圖2.4TD+溫度與震盪頻率關係16 
圖2.5 CFDA認證流程表17 
圖3.1像素亮度階層18 
圖3.2像素矩陣18 
圖3.3Facenet模型結構19 
圖3.4Triple Loss示意圖20 
圖3.5YoloV3 架構圖20 
圖3.6 loU示意圖21 
圖3.7Deep sort流程圖21 
圖3.8動態環境追蹤流程圖22 
圖3.9受測者A動態環境追蹤實驗結果(Zone 3公司-座位) 23 
圖3.10個人動態環境影像追蹤與生理資訊整合結果設計圖24 
圖3.11受測者A動態環境追蹤與生理資訊整合(Zone 3 公司-座位) 25 
圖4.1中繼邊緣平台系統(Edge Gateway)系統架構27 
圖4.2中繼邊緣平台系統(Edge Gateway)現場實測位置圖28 
圖4.3中繼邊緣平台系統天線實拍29 
圖4.4中繼邊緣平台系統S11量測天線的效應29 
圖5.1增強型-中繼邊緣平台系統31 
圖5.2簡易型-中繼邊緣平台系統31 
圖5.3Zone 1 公司-大會議室32 
圖5.4Zone 2 公司-實驗室32 
圖5.5Zone 3 公司-座位32 
圖5.6Zone 4 體育場32 
圖5.7Zone 5 家-餐廳32 
圖5.8Zone 6 家-書房32 
圖5.9Zone 7 家-客廳33 
圖5.10Zone 8 家-房間33 
圖5.11簡易型-中繼邊緣平台系統擺放位置與接收方式33 
圖5.12生理資訊實測結果(Zone 3 公司-座位) 34 
圖5.13生理資訊實測結果(Zone 2 公司-實驗室) 35 
圖5.14生理資訊實測結果(Zone 7 家-客廳) 35 
圖5.15中繼系統配對感測器實測36 
圖5.16中繼系統記錄實測(Zone 3 公司-座位) 36 
圖5.17實測紀錄使用者的Heart rate (Zone 3 公司-座位) 37 
圖6.1運動動態血壓之實測步驟流程圖38 
圖6.2運動動態血壓之預警分析流程圖40 
圖6.3受測者7(血壓無異常變化)之QT實測圖43 
圖6.4受測者7 (血壓無異常變化)之QTc斜率實測圖43 
圖6.5受測者12(血壓異常變化)之QT實測圖44 
圖6.6受測者12(血壓異常變化)之QTc斜率實測圖44 
圖6.7受測者8(血壓無異常變化)之運動後ECG波形45 
圖6.8受測者12(血壓異常變化)之運動後ECG波形46 
圖6.9受測者15(血壓異常變化)之運動後ECG波形46 
圖6.10 HRV預警分析實測圖48 
圖7.1運動血壓風險預警系統設計59 
圖7.2動態環境影像追蹤與生理資訊整合系統59 
圖8.1運動中即時危險預警流程圖60 

 表目錄
表1.1ECG模組規格6 
表1.2先前研究技術比較-本論文ECG模組VS. Polar H78 
表1.3先前研究技術綜合比較9 
表1.4 HRV量測時域變數表11 
表1.5 HRV量測頻域分析變數表12 
表2.1TD+校正溫度電阻與頻率關係表16 
表4.1中繼邊緣平台系統規格26 
表6.1受測者生理基本資訊與測試後臨床表現(受測者1-13) 41 
表6.2受測者生理基本資訊與測試後臨床表現(受測者14-26) 42 
表6.3 SDNN與年齡性別關係47 
表6.4HRV-SDNN受測者同一天同一環境之平均值49 
表6.5HRV-LF/HF受測者同一天同一環境之平均值49 
表6.6 HRV-SDNN受測者同一天同一環境之平均值50 
表6.7HRV-LF/HF受測者同一天同一環境之平均值50 
表6.8HRV-SDNN受測者同一天同一環境之平均值51 
表6.9HRV-LF/HF受測者同一天同一環境之平均值51 
表6.10血壓受測者同一天同一環境之值52 
表6.11血壓受測者同一天同一環境之值52 
表6.12血糖受測者同一天同一環境之值52 
表6.13體溫受測者同一天同一環境之平均值53 
表6.14體溫受測者同一天同一環境之平均值53 
表6.15體溫受測者同一天同一環境之平均值53 
表6.16HRV-SDNN受測者同一天同一環境之平均值54 
表6.17HRV-LF/HF受測者同一天同一環境之平均值54 
表6.18HRV-SDNN受測者同一天同一環境之平均值55 
表6.19HRV-LF/HF受測者同一天同一環境之平均值55 
表6.20血壓受測者同一天同一環境之值56 
表6.21血糖受測者同一天同一環境之值56 
表6.22體溫受測者同一天同一環境之平均值56 
表6.23體溫受測者同一天同一環境之平均值57 
表6.24受測者A的HRV-LF/HF在各區域比較表57

[1]	羅鷹瑞， 楊清源 "心臟疾病資料庫 (第二版)", 中文文學出版社
[2]	ECG, URL: http://www.medicine-on-line.com/html/ecg/e0001ct.htm
[3]	O. Ekwe, A. E. Abioye, Oluwe, K.C. Okoro, "Effective Fading Reduction Techniques in Wireless Communication System", IOSR Journal of Electronics and Communication Engineering (IOSR-JECE),2014, vol. 9, pp. 35-43.
[4]	Y. Ngoko, C. Cérin, "An edge computing platform for the detection of acoustic events", 2017 IEEE International Conference on Edge Computing (EDGE), June 2017, pp. 240-243.
[5]	F. Schroff, D. Kalenichenko, J. Philbin, "FaceNet: A unified embedding for face recognition and clustering", 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR),2015, pp. 815-823.
[6]	W. Lan, J. Dang, Y. Wang, and S. Wang, “Pedestrian detection based on YOLO network model” in Proc. IEEE Int. Conf. Mechatronics Automat., Aug. 2018, pp. 1547–1551.
[7]	A. P. Plageras, K. E. Psannis, C. Stergiou, H. Wang, B. B. Gupta, "Efficient IoT-based sensor BIG data collection-processing and analysis in smart buildings", Future Generation Comput. Syst., Oct. 2017 ,vol. 82, pp. 349-357.
[8]	A. E. Boualouache, O. Nouali, S. Moussaoui, A. Derder, "A BLE-based data collection system for IoT", Proc. 1st Int. Conf. New Technol. Inf. Commun., 2015, pp. 1-5.
[9]	Xin Zhu, Yoshida, K., Yamanobe, W., Yamamoto, Y., Wenxi Chen, Daming Wei, “Conversion of the ambulatory ECG to the standard 12-lead ECG: a preliminary study”, Conference on Biomedical Engineering, 2003. IEEE EMBS Asian-Pacific, 2003, pp. 48-49
[10]	Ihara, Z., Jacquemet, V., Vesin, J.-M., van Oosterom, A., “Adaption of the standard 12-lead ECG system focusing on atrial electrical activity ”, Computers in Cardiology, 2005, pp. 203-205
[11]	呂嘉陞, The Only EKG Book You'll Ever Need, 合記出版社2008
[12]	ECG波形: https://zh.wikipedia.org/wiki/%E5%BF%83%E7%94%B5%E5%9B%BE

[13]	Malik M, Batchvarov VN. "Measurement, interpretation and clinical potential of QT dispersion". J Am Coll Cardiol 2000; 36: 1749–66
[14]	Kim MW, Kim BS, Shin HS, Kim BJ, Sung KC, Kang JH, et al. "Lack of correlation between QTc dispersion and morning blood pressure surge in recently diagnosed essential hypertensive patients". Circ J 2007; 71: 1288-1292.
[15]	Rautaharju PM, Zhou SH, Wong S, Calhoun HP, Berenson GS, Prineas T et al. "Sex differences in the evolution of the electrocardiographic QT interval with age" , Can J Cardiol 1992; 8: 690–5
[16]	S. Cerutti, A. M. Bianchi, L. T. Mainardi, "Spectral Analysis of the Heart Rate Variability Signal", Heart Rate Variability, eds. M. Malik and A. Camm,J. Armonk, NY: Futura Publishing Company, Inc. 1995 p. 63-74.
[17]	Aysin B, Aysin E, "Effect of respiration in heart rate variability (HRV) analysis",  Conf Proc IEEE Eng Med Biol ,2006, Soc 1:1776–1779
[18]	Task Force of The European Society of Cardiology and The North American Society of Pacing and Electrophysiology(Membership of the Task Force listed in the Appendix), “[Guidelines] Heart rate variability-Standards of measurement, physiological interpretation, and clinical use”, European Heart Journal, Vol. 17, pp. 354-381, 1996.
[19]	M. P. Tarvainen, J. P. Niskanen, J. A. Lipponen, P. O. Ranta-Aho, P. A Karjalainen, "Kubios HRV - Heart rate variability analysis software", Computer Methods and Programs in Biomedicine, Jan. 2014, vol. 113, pp. 210-220,.
[20]	CFDA, URL:https://www.cmde.org.cn/CL0122/
[21]	CFDA,URL:http://www.caivd.org/ArticleDetail?combId=efad0775-27b0-4488-86d1-ef9daf7cb660%2Cefad0775-27b0-4488-86d1-ef9daf7cb660&articleId=fa2e9696-3c7a-44e8-9f20-f8c3907b53c3
[22]	Facenet,URL:https://makerpro.cc/2018/12/introduction-to-face-recognition-model-google-facenet/
[23]	http://www.csie.ntnu.edu.tw/~u91029/Image.html
[24]	https://zhuanlan.zhihu.com/p/24837264
[25]	W. Yang, Z. Jiachun, "Real-time face detection based on YOLO", IEEE International Conference on Knowledge Innovation and Invention (ICKII)2018, pp. 221-224.
[26]	Najibi, M., Samangouei, P., Chellappa, R., et al.: "SSH: Single stage headless face detector", In: IEEE International Conference on Computer Vision, IEEE Computer Society (2017), pp. 4885–4894.
[27]	https://mropengate.blogspot.com/2018/06/yolo-yolov3.html
[28]	https://www.itread01.com/content/1545102482.html
[29]	https://www.bluetooth.com/zh-cn/specifications/protocol-specifications/
[30]	P. M. Rautaharju, B. Surawicz, L. S. Gettes et al., “AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part IV: the ST segment, T and U waves, and the QT interval: a scientifc statement from the american heart association electrocardiography and arrhythmias committee, council on clinical cardiology; the american college of cardiology foundation; and the heart rhythm society. endorsed by the international society for computerized electrocardiology”, Journal of the American College of Cardiology 2009, vol. 53, no. 11, pp. 982–991.
[31]	L. Oikarinen, M. S. Nieminen, L. Toivonen et al., “Relation of QT interval and QT dispersion to regression of echocardiographic and electrocardiographic lef ventricular hypertrophy in hypertensive patients: the Losartan Intervention For Endpoint Reduction (LIFE) study”, American Heart Journal 2003, vol. 145, no. 5, pp. 919–925.
[32]	I. Mozos and C. Serban, “Te relation between QT interval and T-wave variables in hypertensive patients”, Journal of Pharmacy and Bioallied Sciences 2011, vol. 3, no. 3, pp. 339–344.
[33]	Samir Saba, MD, Rodolfo Pike, RN, BN, MN, NP, Jennifer Williams, RN, BN, NP, Fadi Khraim, RN, NP, PhD," Arterial Stiffness Is Associated With QTc Interval Prolongation in Patients With Heart Failure", Sage Journals 2018,Vol 20, pp. 255-263 Issue 3.
[34]	XU H, CHEN J, ZHAO Q, et al. "Ambient air pollution is associated with cardiac repolarization abnormalities in healthy adults", [J]. Environ Res, 2019, 171:239-246.
[35]	Yeo TJ, Sharma S."Using the 12-lead electrocardiogram in the care of athletic patients",  Cardiol Clin 2016;34(4):543–55
[36]	Kaltman JR, Thompson PD, Lantos J, Berul CI, Botkin J, Cohen JT, et al. "Screening for sudden cardiac death in the young: report from a National Heart, Lung, and Blood Institute working group", Circulation 2011;123(17):1911–8.
[37]	Marfella R, Gualdiero P, Siniscalchi M, Carusone C, Verza M, Marzano S, Esposito K, Giugliano D." Morning blood pressure peak, QT intervals, and sympathetic activity in hypertensive patients", Hypertension. 2003; 41: 237–243
[38]	Pappachan JM, Sebastian J, Bino BC, Jayaprakash K, Vijayakumar K, Sujathan P, Adinegara LA. "Cardiac autonomic neuropathy in diabetes mellitus: prevalence, risk factors and utility of corrected QT interval in the ECG for its diagnosis", Postgrad Med J. 2008; 84:205–210
[39]	Yi G, Elliott P, McKenna WJ, Prasad K, Sharma S, Guo XH, Camm AJ, Malik M. "QT dispersion and risk factors for sudden cardiac death in patients with hypertrophic cardiomyopathy",  Am J Cardiol.1998; 82:1514–1519
[40]	Straus SM, Kors JA, De Bruin ML, van der Hooft CS, Hofman A, Heeringa J, Deckers JW, Kingma JH, Sturkenboom MC, Stricker BH, Witteman JC." Prolonged QTC interval and risk of sudden cardiac death in a population of older adults", J Am Coll Cardiol. 2006; 47:362–367
[41]	Day CP, McComb JM, Campbell RWF. "QT dispersion: an indicator of arrhythmia risk in patients with long QT intervals", Br Heart J 1990; 63: 342–4.
[42]	https://www.joiiup.com/knowledge/content/1460