臺灣博碩士論文加值系統

隨著醫療科技與生活水平的上升，人們的平均壽命也跟著上升，根據中華民國國家發展委員會對於臺灣的人口結構統計推估，在2018年臺灣正式邁入了高齡社會。而根據聯合國世界衛生組織的定義65歲以上的人口占總人口14%以上則邁入「高齡社會」。為了照護這些老年人，政府提出了長照計畫，而長照指的是對身心功能障礙者，在一段長時間內，提供一套包括長期性的醫療、護理、個人、與社會支持的照顧，其目的在促進或維持身體功能，增進獨立自主的正常生活能力。
根據內政部統計處的統計，在107年12月全臺灣人口共有23,588,932人，依照上訴所說全臺灣的老年人人口約有330萬人以上。依據衛生福利部統計處的統計，在老人長期照顧、安養機構的工作人員共有25,733人，其中護理人員有5,211人，在一般護理之家的護理人員及照顧服務員共有15,777人，護理人員有4,304人。根據上面的數據顯示平均1位醫護人員要照顧約347位老年人，就算扣除有家庭照顧的老人，1位醫護人員還是要照顧約200~300位老年人，隨著時間的不斷的往後推移老年人會越來越多，醫護人員的人力會越來越不夠。因為人力的不足，有可能導致醫護人員的照顧疏失，照顧疏失則有可能造成老年人的生命危險，像跌倒後的骨折、頭部碰撞尖銳物品導致腦出血…等傷害，嚴重一些或長時間無人發現的話，可能導致老年人死亡。
為了解決這些問題，本研究提出了使用Microsoft Kinect感測器所提供的骨架訊息來偵測跌倒事件的發生，在被偵測者開始往下降時進行判斷，當系統判斷被偵測者發生跌倒事件後則會發送警訊訊息給相關醫護人員進行提醒，讓醫護人員能第一時間到跌倒事件發生地點進行急救以保住被偵測者的生命。
最後經過數據的收集與分析並把分析後的數據套用到本研究的系統上之後實測發現，本研究的系統偵測到發生跌倒事件的準確率高於90%，被誤偵測倒發生跌倒事件的機率低於1%。

With the advancement in medical technology and the standards of living, the average life expectancy has also been increased. According to the extrapolation that the National Development Council conducted on Taiwan's demographic statistics, it is indicated that Taiwan is officially stepping into the Aged society since 2018. As the definition of the Aged society of WHO, when the proportion of the population in a society that is comprised of persons at age 65 or older exceeds 14%, it is an "aged society." For the purpose of providing care services to these the elderly people, the government has promoted a project named Long-term Care Plan, which is aimed to service the elderly with disabilities. Within a long term, the government provides a series of long-term care services such as medical treatment, nursing care and supports for personal care and social supports. The purposes of promoting the plan are to improve or maintain physical abilities and thus to increase the capability of independent living.
According to the statistics issued in December of 2018 by the Department of Statistics, the Ministry of the Interior, there are 23,588,932 people living in Taiwan, and the population of the elderly in Taiwan is more than 3.3 million or so as mentioned above. And the statistics made by the Ministry of Health and Welfare indicate, the employees working in nursing homes and institutions of long-term care for the elderly are 25,733 persons in total and the number of nurse practitioners is 5,211 which has been included. In general nursing homes, there are 15,777 nurse practitioners and care attendants, and the number of nurse practitioners is 4,304. As the data indicated above, the number of the elderly person that one healthcare worker takes care with are 347 persons on average and the number is decreased to 200-300 elderly persons if excluding the elderly persons who were taken care with the families. As the time goes by, the number of the elderly is increasing, and the number of healthcare workers will be in a serious shortage accordingly. Due to the insufficient health-care workers, there may be chances to lead caring negligence which may possibly causes threats to danger the life such as the hemorrhage caused by the bone fracture after falling down or hitting the head by sharpen objects and so on. In some severe cases, the elderly may die due to the worse issues or not being found for a long period.
In order to solve these issues, this study proposes the use of Sensor of Microsoft Kinect for detecting the occurrence of falling down with the skeleton status of the subject. The sensor is started to process and judge when the subject was falling. When the event of falling down is detected and filed, the system will send an alert message to inform the related healthcare workers and they can provide first aids to save the life or provide medical treatments in time if necessary when arrived at the soonest speed.
In the end, after analyzing the collected data and performing the actual testing by applying it to the system operated in the study, it is found that the precision of detectable occurrence of falling down is higher than 90%, and the probability of false detection is less than 1%.

封面內頁
簽名頁
中文摘要 iii
Abstract v
誌謝 vii
目錄 viii
圖目錄 x
表目錄 xii

第一章 前言 1
1.1研究背景 1
1.2研究動機與目的 1
1.3論文架構 3
第二章 相關研究 4
2.1穿戴式感測器 4
2.2影像式感測器 7
2.3體感感測器 10
第三章 研究方法 13
3.1骨架節點定義 13
3.2偵測跌倒的研究方法 14
3.2.1判斷前後移動或者左右移動 15
3.2.2偵測倒向左或右方向 16
3.2.3偵測倒向前或後方向 18
3.3 數據收集與分析 20
第四章 研究成果與實作展示 30
4.1研究成果 30
4.2偵測準確度 32
第五章 結論與未來展望 38
參考資料 39

[1] 中華民國國家發展委員會（2018年）。取自
https://www.ndc.gov.tw/Content_List.aspx?n=695E69E28C6AC7F3。
[2] 趙明玲，方郁文，高淑芬（2005年）。社區老年人跌倒之預防及護理措施。財團法人聯新文教基金會，桃園市。
[3] 衛生福利部國民健康署（2015年）。取自
https://www.hpa.gov.tw/Home/Index.aspx。
[4] 蔡益堅（2012年）。老人跌倒的因素與防制策略，家庭計畫通訊第八卷_第142期。衛生福利部國民健康署，台北市。
[5] 周尚賢（2016年）。整合Kinect與Unity3D體感應用之研究。國立屏東大學碩士論文，屏東縣。取自https://hdl.handle.net/11296/vtwd8z。
[6] J. Y. Hwang, J. M. Kang, Y. W. Jang, and H. C. Kim, “Development of Novel Algorithm and Real-time Monitoring Ambulatory System Using Bluetooth Module for Fall Detection in the Elderly,” in 26th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 2204-2207, 2004.
[7] Chin-Feng Lai, Sung-Yen Chang, Han-Chieh Chao, Yueh-Min Huang, “Detection of Cognitive Injured Body Region Using Multiple Triaxial Accelerometers for Elderly Falling”, in IEEE Sensors Journal, Vol.11, Issue.3, 2011, pp. 763-770.
[8] Natthapon Pannurat, Surapa Thiemjarus, Ekawit Nantajeewarawat, “A Hybrid Temporal Reasoning Framework for Fall Monitoring”, in IEEE Sensors Journal, Vol.17, Issue.6, 2017, pp.1749-4759.
[9] Saminda W. Abeyruwan, Dilip Sarkar, Faisal Sikder, Ubbo Visser, “Semi-Automatic Extraction of Training Examples From Sensor Readings for Fall Detection and Posture Monitoring”, in IEEE Sensors Journal, Vol.16, Issue.13, 2016, pp. 5406-5415.
[10] Lina Tong, Quanjun Song, Yunjian Ge, Ming Liu, “HMM-Based Human Fall Detection and Prediction Method Using Tri-Axial Accelerometer”, in IEEE Sensors Journal, Vol.13, Issue.5, 2013, pp. 1849-1856.
[11] Jian Liu, Thurmon E. Lockhart, “Development and Evaluation of a Prior-to-Impact Fall Event Detection Algorithm”, in IEEE Transactions on Biomedical Engineering, Vol.61, Issue.7, 2014, pp. 2135-2140.
[12] Andreas Ejupi, Matthew Brodie, Stephen R. Lord, Janneke Annegarn, Stephen J. Redmond, Kim Delbaere, “Wavelet-Based Sit-To-Stand Detection and Assessment of Fall Risk in Older People Using a Wearable Pendant Device”, in IEEE Transactions on Biomedical Engineering, Vol.64, Issue.7, 2017, pp. 1602-1607.
[13] F. Sposaro and G. Tyson, “iFall: An android application for fall monitoring and response,” in 2009 Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 6119-6122, 2009
[14] Damien Brulin, Yannick Benezeth, Estelle Courtial, “Posture Recognition Based on Fuzzy Logic for Home Monitoring of the Elderly”, in IEEE Transactions on Information Technology in Biomedicine, Vol.16, Issue.5, 2012, pp. 974-982.
[15] P. Viola, M.Jones, “Rapid Object Detection using a Boosted Cascade of Simple”, in IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Issue.1,2001, pp.511-518.
[16] Edouard Auvinet, Franck Multon, Alain Saint-Arnaud, Jacqueline Rousseau, Jean Meunier, “Fall Detection With Multiple Cameras: An Occlusion-Resistant Method Based on 3-D Silhouette Vertical Distribution”, in IEEE Transactions on Information Technology in Biomedicine, Vol.15, Issue.2, 2011, pp. 290-300.
[17] Diogo Ferreira , Raul Oliveira , Octavian Postolache，“Physical Rehabilitation based on Kinect Serious Games”，2017 Eleventh International Conference on Sensing Technology （ICST），2017.
[18] Shihab Shahriar Hazari, Asaduzzaman, Lamia Alam and Nasim Al Goni，“Designing a Sign Language Translation System using Kinect Motion Sensor Device”，International Conference on Electrical, Computer and Communication Engineering （ECCE）, February 16-18, 2017