本論文設計一個訓練系統以幫助照護者在學習搬移臥床病人過程中，避免因姿勢不正確而受傷。 病人搬移過程中，最主要動作為將病人從病床上轉移到輪椅。 整個系統的回饋將協助護理老師觀察照顧者的表現，尤其是姿勢。此系統硬體為一台電腦與一台三維深度感測器。深度感測器用以分別抓取護理老師的示範標準動作與照顧者的動作影像。透過系統，取得護理老師與照顧者的人體骨架，並比較彼此的相似度。

現有由三維深度感測器得到影像資料並計算人體骨架的套裝程式，有一個缺點:當兩個人太接近時或被遮蔽時會導致人體骨架不準確或是轉到另一個人。本研究稱此為人體姿勢遮蔽問題。因為有此問題，現有套裝軟體不適用本系統。因此本研究引進卡爾曼濾波器，以解決此問題。卡爾曼濾波器被應用在追蹤人體骨架的每一個關節點，並預估可能被遮蔽時可能的位置。最後結果顯示卡爾曼濾波器確實改善了大部分現有深度感測器配合的套裝軟體的人體姿勢遮蔽問題。

This thesis is to design a training system for assisting nursing students or caregivers in learning transfer skills to avoid injury. The system is focused on transferring a patient lying on bed to a wheelchair. A nursing teacher would use the proposed system to help nursing students in checking their performance, especially postures. The hardware platform is a personal computer equipped with a depth sensor. The depth sensor is used for capturing the images and postures with skeletons of the teacher and the caregivers.

Currently available depth sensor software packages generally cannot handle the body overlap issue. There are two bodies and their bodies are very close, the package would generate wrong postures for the actors. Hence it is very difficult to separate and track the bodies clearly of the caregiver and patient who are close interacting during the patient transfer process. To solve this issue, Kalman filters are introduced in this research to track every joint and predict its position when it is hidden. The experimental results showed that Kalman filter improves most of the body overlap cases.

[1] A. Garg, B. D. Oern, and B. Carlson, ”An ergonomic evaluation of nursing assistants’job in a nursing home” Ergonomics, vol. 35, pp. 979-995, Sep. 1992.
[2] L. Liu, and S. Mehrotra, ”Patient Associated Motion Detection with Optical Flow using Microsoft Kinect v2”, IEEE/ACM International Conference on Connected Health: Applications, Systems and Engineering Technologies (CHASE), PP. 274-275, 2017.
[3] M. Capecci, M. G. Ceravolo, F. Ferracuti, S. Iarlori, V. Kyrki, S. Longhi, L. Romeo, and F. Verdini, ”Physical Rehabilitation Exercises Assessment based on Hidden Semi-Markov Model by Kinect v2”, IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI), pp. 256-259, 2016.
[4] H. T. Chen, Y. Z. He, C. L. Chou, S. Y. Lee, B. S. P. Lin, and J. Y. Yu, ”Computerassisted Self-training System for Sports Exercise using Kinects”, IEEE International Conference on Multimedia and Expo Workshops (ICMEW), pp. 1-4, 2013.
[5] C. Hoang, H. T. Dang, and V. D. Nguyen, ”Kinect-based Virtual Training System for Rehabilitation”, International Conference on System Science and Engineering (ICSSE), Ho Chi Minh City, Vietnam, pp. 53-56, 2017.
[6] W. Zhao, Q. Wu, D. D. Espy, M. A. Reinthal, X. Luo, and Y. Peng, ”A Feasibility Study on using a Kinect-based Human Motion Tracking System to Promote Safe Patient Handling”, IEEE International Conference on Electro Information Technology (EIT), pp. 462-466, 2017.
[7] S. Gauthier, and A. M. Cretu, ”Human Movement Quantification using Kinect for In-Home Physical Exercise Monitoring”, IEEE International Conference on Computational Intelligence and Virtual Environment Systems and Applications (CIVEMSA), pp. 6-11, 2014.
[8] W. Zhao, M. A. Reinthal, D. D. Espy, and X. Luo, ”Rule-Based Human Motion Tracking for Rehabilitation Exercises: Realtime Assessment, Feedback, and Guidance”, IEEE Access, Special Section on Intelligent Systems for the Internet of Things, Oct., pp. 21382-21394, 2017.
[9] Z. Huang, A. Nagata, M. K. Pak, J. Maeda, Y. Kitajima, M. Nakamura, K. Aida, and N. Kuwahara, ”Self-Help Training System for Nursing Students to Learn Patient Transfer Skills”, IEEE Transactions on Learning Technologies, Vol. 7, No. 4, Oct., pp. 319-332, 2014.
[10] Z. Huang, C. Lin, M. K. Pak, J. Maeda, Y. Kitajima, M. Nakamura, N. Kuwahara, T. Ogata, and J. Ota, ”Robot Patient Design to Simulate Various Patients for Transfer Training”, IEEE/ASME Transactions on Mechatronics, Vol. 22, No. 5, Oct., pp. 2079-2090, 2017.
[11] Z. Huang, C. Lin, M. K. Pak, J. Maeda, Y. Kitajima, M. Nakamura, N. Kuwahara, T. Ogata, and J. Ota, ”Impact of Using a Robot Patient for Nursing Skill Training in Patient Transfer”, IEEE Transactions on Learning Technologies, Vol. 10, No. 3, July, pp. 355-366, 2017.
[12] M. S. Arulampalam, S. M. Gordon, and T. Clapp, ”A Tutorial on Particle Filters for Online Nonlinear/Non-Gaussian Bayesian Tracking”, IEEE Transactions on Signal Processing, Vol. 50, No. 2, Feb., pp. 174-188, 2002.
[13] G. Welch, and G. Bishop, ”An introduction to the Kalman Filter”, TR 95-041, Department of Computer Science, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-3175, 2006, Available on Association for Computing Machinery (ACM) Digital Library: https://dl.acm.org/citation.cfm?id=897831.
[14] PrimeSense, The official document: Prime Sensor NiTE 1.3 Algorithms notes, version 1.0.
[15] L. W. Kang, C. S. Lu, and C. Y. Hsu, ”Compressive Sensing-based Image Hashing”, IEEE International Conference on Image Processing (ICIP), Cairo, Egypt, pp. 1285-1288, 2009.
[16] Y. J. Chen, An Automatic Assessment Framework for Exercise Training System using Hidden Markov Model and K-Means Clustering, Master Thesis, Department of Engineering Science, National Cheng Kung University (NCKU), Tainan, Taiwan, 2017.
[17] T. L. Cheng, An IoT-Based Exercise Training System, Master Thesis, Department of Engineering Science, National Cheng Kung University (NCKU), Tainan, Taiwan, 2016.