本論文探討利用人形機器人模擬真人動作演出時，在相似度上的評估與可能的改進方法。為了使人形機器人能完整表現出與人類相同的表演動作，我們必須考慮許多可能造成影響的因素。除了在人類運動關節可能與機器人關節數目不盡相同之外，尚有關節的轉動角度、角速度限制，肢段長度不同造成動作角度不一致的情況，另外如何維持平衡、如何使動作更相近於原本所要求的動作…等等，更是一個人形機器人要完成各項動作表演的基本要求。如何在有限制的情況之下，讓人形機器人能夠做出與真人相同的表演動作，即使無法完全相同，仍能以合理的方式來評量相似程度，進而來修正動作，達到表演的完整，是本論文的研究方向。本論文利用架構對應區塊法，將由動作擷取系統上所採集下來的動作對應至利用Webots機器人模擬軟體所建立的3D機器人模型上。為了評估機器人演示的動作與原始人體動作的相似程度，本論文提出了關鍵動作達成率的概念，並實作架構對應區塊法中的N-1與1-1對應，嘗試將機器人可動範圍發揮至極限來使動作調整至合理的動作角度並兼顧表演動作的完整性。實驗的結果顯示，透過架構對應區塊法並輔以關鍵動作達成率來修正後的結果，結合1-1對應與N-1對應，考量關節角度與肢段長度差異程度的情況之下，比僅使用1-1對應，僅考量關節角度的情形，能產生出更為接近原始動作的轉譯。

This thesis is a collection of work on programming humanoid robots to mimic human motions. There are lots of factors to consider when we want to evaluate and even improve the similarity of the motions performed by humanoid robots comparing to the original human motions. For example, the skeleton, the joint numbers, the angles of joints, the limit of rotational speed, and the body segmental of every humanoid robot may be all or partly different from human body. Besides these, how to keep the robot balanced is also an important issue. In this thesis, an evaluation criterion named “key pose completion ratio” is used to measure how a programmed humanoid motion is close to the original motion. Also an approach which is based on the “structural mapping zone” is used to translate human motions into robot motions. We use motion capture system to collect human motions, and we use Webots, a 3D model based robot simulator, to demonstrate the translated robot motions. To find the better solution, we implement both 1-1 mapping but N-1 mapping of structural mapping zone approach to complete the translation. By comparing the two versions of robot motions with their key pose completion ratios, we can clearly observe that N-1 mapping gives more similar motion than 1-1 mapping.

[1] Humanoid Robotics Institute Waseda University 網站
http://www.humanoid.waseda.ac.jp/index.html
[2] Hirai, K., Hirose, M., Haikawa, Y., and Takenaka, T., “The Development of Honda Humanoid Robot,” in Proceedings of the IEEE International Conference on Robotics & Automation, Vol.2, pp. 1321-1326, 16-20 May 1998.
[3] Sakagami, Y., Watanabe, R., Aoyama, C., Matsunaga, S., Higaki, N., and Fujimura, K., “The intelligent ASIMO: system overview and integration,” in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and System, vol. 3, pp. 2478-2483, 2002.
[4] Arakawa, T. and Fukuda, T., “Natural motion trajectory generation of biped locomotion robotusing genetic algorithm through energy optimization,” in Proceedings of the IEEE International Conference on Systems, Man, and Cybernetic, vol. 2, pp. 1495-1500, 14-17 Oct 1996.
[5] Nakaoka, S., Nakazawat, A., Yokoit, K., and Ikeuchi, K., “Leg Motion Primitives for a Dancing Humanoid Robot,” in Proceedings of the IEEE International Conference on Robotics and Automation ICRA ’04, vol. 1, pp. 610-615, 26 April – 1 May 2004.
[6] Geppert, L., “QRIO: The Robot That Could. A technical marvel, Sony's biped humanoid promises to revolutionize home entertainment,” IEEE Spectrum, vol. 41, issue 5, pp. 34-37, 2004.
[7] 洪志憲，<利用架構對應區塊之人形機器人動作生成方法>，國立台灣科技大學資訊工程系碩士論文，2007年7月。
[8] 晉茂林，<機器人學>，五南圖書公司，台北，2000。
[9] Pollard, N. S., Hodgins, J. K., Riley, M. J., and Atkeson C. G., “Adapting human motion for the control of a humanoid robot,” in Proceedings of the IEEE International Conference on Robotics and Automation ICRA '02, vol. 2, pp. 1390-1397, 2002.
[10] Matsui, D., Minato, T., MacDorman, K. F., and Ishiguro, H., “Generating Natural Motion in an Android by Mapping Human Motion,” in Proceedings of the 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 3301-3308, 2-6 Aug. 2005.
[11] Ko, H. and Badler, N. I., “Animating human locomotion with inverse dynamics,” in Proceedings of the Computer Graphics and Applications, IEEE, vol. 16, Issue 2, pp.50-59, Mar 1996.
[12] Vicon 官方網站
http://www.vicon.com/
[13] Webots官方網站
http://www.cyberbotics.com/