隨著運動追蹤系統的進步，其產物運動追蹤資料被廣泛的應用於各種領域，本研究首先介紹運動資料儲存方式，藉由圖像技術對運動數據進行編碼，將原始文字檔形式的資料，以顏色圖檔呈現，並將檔案壓縮。
在運動追蹤結束後，運動資料平順是一項重要處理程序，藉由補點技術可將因遺失標誌框而錯誤紀錄的數據修正，接著利用平滑化程序，可將包含於資料中的雜訊濾除；本研究將利用模型的優點開發策略，藉此精準的濾除錯誤部分，以完整保留運動資訊的特性。
最後，利用平順程序處理後的運動資料，尋找運動曲線過零點位置，藉由分析資料的過零點個數，發展一自動化運動資料分割技術，此技術可將一完整的運動資料切割成片段，此目的是要萃取出運動的招式，以利於運動比較的應用。

With the advancement of motion system, the product, motion data, has widely used in a lot of fields. For the first part, this study introduces the format of storage for motion data. By the method of motion data coding, we can get a BMP file. This files present by gray or color instead of text. It will provide an intuitive value change and compress archive.
For the second part, this study discuss the fairing process. After Mocap experiments, the motion data is needed to do fairing in order to correct the wrong data. In this study, there are four parts for motion data fairing including filtering, interpolation, rectification and smoothing. Filtering and interpolation is used to deal with the missing marker problem. Rectification and smoothing is used to deal with the high-frequency noise. This study will develop some strategies for filtering by the advantages of the model. It can accurately filter off the wrong value.
After the fairing process, we can get the smooth motion data and do the motion data segmentation. The concept of segmentation is the analysis of extreme value on motion curve. Extreme value on motion curve is also called the zero-crossing point because the slope is zero. This study will analyze the number of zero-crossing points to develop a method of automatic motion data segmentation. This method can separate a complete motion data into several motion primitives. The motion primitives can be used for motion comparison.

[1] Loy, Gareth, Josephine Sullivan, and Stefan Carlsson. "Pose-based clustering in action sequences." Higher-Level Knowledge in 3D Modeling and Motion Analysis, 2003. HLK 2003. First IEEE International Workshop on. IEEE, 2003.
[2] Kadone, Hideki, and Yoshihiko Nakamura. "Segmentation, memorization, recognition and abstraction of humanoid motions based on correlations and associative memory." Humanoid Robots, 2006 6th IEEE-RAS International Conference on. IEEE, 2006.
[3] Mori, Takaki, and Kuniaki Uehara. "Extraction of primitive motion and discovery of association rules from motion data." Robot and Human Interactive Communication, 2001. Proceedings. 10th IEEE International Workshop on. IEEE, 2001.
[4] Barbič, Jernej, et al. "Segmenting motion capture data into distinct behaviors."Proceedings of the 2004 Graphics Interface Conference. Canadian Human-Computer Communications Society, 2004.
[5] Gong, Yihong, and Xin Liu. "Video summarization using singular value decomposition." Computer Vision and Pattern Recognition, 2000. Proceedings. IEEE Conference on. Vol. 2. IEEE, 2000.
[6] Cooper, Matthew, and Jonathan Foote. "Summarizing video usingnon-negative similarity matrix factorization." Multimedia Signal Processing, 2002 IEEE Workshop on. IEEE, 2002.
[7] Bulut, Eyuphan, and Tolga Capin. "Key frame extraction from motion capture data by curve saliency." Computer Animation and Social Agents. 2007.
[8] Bindiganavale, Rama, and Norman I. Badler. "Motion abstraction and mapping with spatial constraints." Modelling and Motion Capture Techniques for Virtual Environments. Springer Berlin Heidelberg, 1998. 70-82.
[9] Schulz, Sebastian, and Annika Woerner. "Automatic motion segmentation for human motion synthesis." Articulated Motion and Deformable Objects. Springer Berlin Heidelberg, 2010. 182-191.
[10] Shiratori, Takaaki, Atsushi Nakazawa, and Katsushi Ikeuchi. "Detecting dance motion structure through music analysis." Automatic Face and Gesture Recognition, 2004. Proceedings. Sixth IEEE International Conference on. IEEE, 2004.
[11] So, Clifford KF, and George Baciu. "Entropy-base motion extraction for motion capture animation. " Computer Animation Virtual World 16.3-4(2005): 225-235.
[12] Fod, Ajo, Maja J. Matarić, and Odest Chadwicke Jenkins. "Automated derivation of primitives for movement classification." Autonomous robots 12.1 (2002): 39-54.
[13] Choensawat, Worawat, Woong Choi, and Kozaburo Hachimura. Similarity Retrieval of Motion Capture Data Based on Derivative eatures." JACIII 16.1 (2012): 13-23.
[14] Kim, Ming-Hwa, Lap-Pui Chau, and Wan-Chi Siu. "Keyframe selection or motion capture using motion activity analysis." Circuits and Systems ISCAS), 2012 IEEE International Symposium on. IEEE, 2012.
[15] Onder, Onur, et al. "Keyframe reduction techniques for motion capture ata."3DTV Conference: The True Vision-Capture, Transmission and Display of 3D Video, 2008. IEEE, 2008.
[16] Guerra-Filho, Gutemberg, Cornelia Fermuller, and Yiannis Aloimonos. "Discovering a language for human activity." Proceedings of the AAAI 2005 fall symposium on anticipatory cognitive embodied systems, Washington, DC. 2005.
[17] 龍學勇, "人體運動處理系統之研究,"國立成功大學機械工程學系博論文, 2013.
[18] 鄭秀瑗, " 現代運動生物力學, 2ed." 國防工業出版社: 國煩工業出社, 2007.
[19] Kavan, Ladislav, et al. "Dual quaternions for rigid transformation lending."Technical report (2006).
[20] Soudan, Karel, and Paul Dierckx. "Calculation of derivatives and ouriercoefficients of human motion data, while using spline unctions." Journal of biomechanics 12.1 (1979): 21-26.