混合實境(Mixed-Reality) 技術的出現造就了許多新的可能性，應用在醫療、娛樂與
教育上的相關產品也越來越多，而特別在教育這一塊，MR 更是因為其特性而能夠帶給使
用者更好的體驗與學習效果。然而目前的MR 產品應用情境也多半侷限在使用者自學的情
境，鮮少有能夠幫助教練抑或是專家檢視其學員狀況的應用。我們所提出的MR 教練輔
助系統正是為了幫助教練能夠更快且更準確地完成檢視作業的系統。不同於傳統平面影片
僅用攝影機拍攝，我們使用了動作捕捉裝置Kinect 作為資料輸入端，將補捉的學員動作
與教練所示範的正確動作做比較分析，得出分析報告並放入我們所使用的頭戴式MR 裝
置HoloLens，最後由教練配戴HoloLens 並實際操作我們的系統。在檢視MR 影片的過程
中，教練可以得到額外的學員動作錯誤提示的視覺化回饋，透過這個機制再搭配三維空間
的點雲資訊能夠讓教練更容易發現學員所作動作之不足之處。

The rise of Mixed-Reality brings many new possibilities to our life. There are also
more and more applications came with MR in the field of medical, entertainments and
education.MR is more capable of bringing not only stunning experiences but also huge
outcome of learning to the users especially in the area of education. However, most of the
existing MR applications are constrained in the function of self-learning.Only very few of
them are aim for coaches or experts to check the status of their students.As a result, we
provide the system that helps coaches by aiding them to finish their task more quickly and
more efficiently.Unlike tradition video clips are taken by the camera, we use motion capture
device like Kinect to record the video as the input of our system.Then we will compare the
motion performed by students with the ground truth motion recorded by coaches. After
that, we will out this report file to the MR device we used which is called HoloLens.
Finally, we will make the coach to wear the HoloLens and use our system. In the process
of checking the status of students. The coach is available to get additional visualization
feedback of the wrong part performed by students.By combining this mechanism and three
dimensional point cloud information will benefit the coach from being easier to find where
is the insufficient part through the motion performed by students.

緒論-1
相關研究-3
系統概觀-5
錄製階段-6
分析階段-8
渲染階段-12
系統驗證-16
結果與討論-21
結論-26
附錄-28

[1] J. C. P. Chan, H. Leung, J. K. T. Tang, and T. Komura. A virtual reality dance training
system using motion capture technology. IEEE Transactions on Learning Technologies, 4
(2):187–195, April 2011.
[2] Ping-Hsuan Han, Yang-Sheng Chen, Yilun Zhong, Han-Lei Wang, and Yi-Ping Hung. My
tai-chi coaches: An augmented-learning tool for practicing tai-chi chuan. In Proceedings of
the 8th Augmented Human International Conference, AH ’17, pages 25:1–25:4, New York,
NY, USA, 2017. ACM. ISBN 978-1-4503-4835-5. doi: 10.1145/3041164.3041194. URL
http://doi.acm.org/10.1145/3041164.3041194.
[3] K. Hachimura, H. Kato, and H. Tamura. A prototype dance training support system with
motion capture and mixed reality technologies. In RO-MAN 2004. 13th IEEE International
Workshop on Robot and Human Interactive Communication (IEEE Catalog No.04TH8759),
pages 217–222, Sept 2004.
[4] Sergio Orts, Christoph Rhemann, Sean Fanello, David Kim, Adarsh Kowdle, Wayne Chang,
Yury Degtyarev, Philip L. Davidson, Sameh Khamis, Minsong Dou, Vladimir Tankovich,
Charles Loop, Qin Cai, Philip Chou, Sarah Mennicken, Julien Valentin, Pushmeet Kohli,
Vivek Pradeep, Shenlong Wang, and Shahram Izadi. Holoportation: Virtual 3d teleportation
in real-time. 10 2016.
[5] Thomas Olsson, Pirita Ihamäki, Else Lagerstam, Leena Ventä-Olkkonen, and Kaisa Väänänen.
User expectations for mobile mixed reality services: An initial user study. In VTT
Symposium (Valtion Teknillinen Tutkimuskeskus), page 19, 01 2009.
[6] Minsu Jang, D. Kim, Y. Kim, and Jaehong Kim. Automated dance motion evaluation
using dynamic time warping and laban movement analysis. In 2017 IEEE International
Conference on Consumer Electronics (ICCE), pages 141–142, Jan 2017. doi: 10.1109/
ICCE.2017.7889261.
[7] Andreas Aristidou, Efstathios Stavrakis, Panayiotis Charalambous, Yiorgos Chrysanthou,
and Stephania Loizidou Himona. Folk dance evaluation using laban movement analysis. J.Comput. Cult. Herit., 8(4):20:1–20:19, August 2015. ISSN 1556-4673. doi: 10.1145/2755566.
URL http://doi.acm.org/10.1145/2755566.
[8] M. Kowalski, J. Naruniec, and M. Daniluk. Livescan3d: A fast and inexpensive 3d data
acquisition system for multiple kinect v2 sensors. In 2015 International Conference on 3D
Vision, pages 318–325, Oct 2015. doi: 10.1109/3DV.2015.43.
[9] F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel,
P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau,
M. Brucher, M. Perrot, and E. Duchesnay. Scikit-learn: Machine learning in Python. Journal
of Machine Learning Research, 12:2825–2830, 2011.