桌球運動與電腦視覺結合的研究近年來蓬勃發展，其中桌球機器人是一套複雜的系統，包含了許多技術，最主要的困難在於如何預判球的路徑與位置。我們利用Kinect取得發球員骨架資訊，進一步考慮球拍與球的動作，以此辨別發球的旋轉方向。並採用支持向量機(SVM)訓練每種發球球路的訓練模型，這些訓練模型用來分類測試集並辨識發球的旋轉類型。為了能夠提升發球辨識率，我們嘗試混和不同的特徵集資料以及調整訓練集大小，比較並探討在不同球員的發球姿勢的發球辨識率。

　　Vision-based sport analysis becomes a popular research topic these years. The most challenging problem of a table tennis robot is to analyze and predict the trajectory of the table tennis ball. We utilize Kinect to capture the skeleton of the table tennis player. Combining with the detection and tracking of table tennis ball and racket, we recognize the spin serve type and predict the trajectory of the table tennis ball. An individual Support Vector Machine (SVM) is adopted to train the model of each spin serve type and then all the models are used to classify the test data to identify the type of the spin serve. We try to mix different set of features and adjust the size of the training set to optimize the accuracy of the serve type recognition. The accuracy of the recognition results on different player’s postures is compared and discussed.

致謝　I
摘要　II
Abstract　III
目錄　IV
圖目錄　VI
表目錄　VII
第一章 序論　1
1.1 研究動機　1
1.2 研究目的　1
1.3 論文概述　2
第二章 文獻探討　3
2.1 軌跡追蹤相關文獻　3
2.2 球旋轉相關文獻　4
2.3 骨節點動作辨識　4
2.3.1 特徵擷取　5
2.3.2 動作辨識　6
第三章 研究方法　9
3.1 系統架構　9
3.2 前處理　10
3.2.1 背景相減法(Background subtraction)　10
3.2.2 Kinect獲取骨節點　10
3.3 球拍、球偵測　11
3.4 特徵擷取　12
3.5 SVM訓練　14
3.6 預測發球動作　15
第四章 實驗結果　17
4.1 實驗設計　17
4.2 實驗環境　17
4.3 實驗訓練資料　18
4.4 發球辨識　18
4.5 3D軌跡重建　26
第五章 結論與未來展望　29
5.1 論文貢獻　29
5.2 未來研究方向　29
5.3 總結　29
參考文獻　31

[1]A. F. Bobick and J. W. Davis, "The recognition of human movement using temporal templates," in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 23, no. 3, pp. 257-267, Mar 2001.
[2]C. Chang and C. Lin, “LIBSVM: A library for support vector machines.” ACM Transactions on Intelligent Systems and Technology, 2:27:1--27:27, 2011. Software available at http://www.csie.ntu.edu.tw/~cjlin/libsvm
[3]C. Fengjiang, D. Muqing and W. Cong, "Kinect-based gait recognition system design via deterministic learning," 2017 29th Chinese Control And Decision Conference (CCDC), Chongqing, 2017, pp. 5916-5921.
[4]H. Myint, P. Wong, L. Dooley and A. Hopgood, "Tracking a table tennis ball for umpiring purposes," 2015 14th IAPR International Conference on Machine Vision Applications (MVA), Tokyo, 2015, pp. 170-173.
[5]J. Feng, B. Price, S. Cohen and S. F. Chang, "Interactive Segmentation on RGBD Images via Cue Selection," 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, 2016, pp. 156-164.
[6]J. Liu, Z. Fang, K. Zhang and M. Tan, "Improved high-speed vision system for table tennis robot," 2014 IEEE International Conference on Mechatronics and Automation, Tianjin, 2014, pp. 652-657.
[7]L. Xia and J. K. Aggarwal, "Spatio-temporal Depth Cuboid Similarity Feature for Activity Recognition Using Depth Camera," 2013 IEEE Conference on Computer Vision and Pattern Recognition, Portland, OR, 2013, pp. 2834-2841.
[8]L. Xia, C. C. Chen and J. K. Aggarwal, "View invariant human action recognition using histograms of 3D joints," 2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, Providence, RI, 2012, pp. 20-27.
[9]M. D. Bengalur, "Human activity recognition using body pose features and support vector machine," 2013 International Conference on Advances in Computing, Communications and Informatics (ICACCI), Mysore, 2013, pp. 1970-1975.
[10]O. Patsadu, C. Nukoolkit and B. Watanapa, "Human gesture recognition using Kinect camera," 2012 Ninth International Conference on Computer Science and Software Engineering (JCSSE), Bangkok, 2012, pp. 28-32.
[11]Q. Wang, K. Zhang and D. Wang, "The trajectory prediction and analysis of spinning ball for a table tennis robot application," The 4th Annual IEEE International Conference on Cyber Technology in Automation, Control and Intelligent, Hong Kong, 2014, pp. 496-501.
[12]R. Mukai, T. Asano and H. Hara, "Analysis and evaluation of tennis plays by computer vision," 2011 IEEE International Conference on Mechatronics and Automation, Beijing, 2011, pp. 784-788.
[13]R. Mukai, T. Asano and H. Hara, "Analysis and evaluation of tennis plays by computer vision," 2011 IEEE International Conference on Mechatronics and Automation, Beijing, 2011, pp. 784-788.
[14]S. Kuo, “Vision Based Analysis for Table Tennis Game,” Master thesis, National Dong Hwa University, Hualien 2008.
[15]S. Tamaki and H. Saito, "Reconstruction of 3D Trajectories for Performance Analysis in Table Tennis," 2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops, Portland, OR, 2013, pp. 1019-1026.
[16]T. L. Le, M. Q. Nguyen and T. T. M. Nguyen, "Human posture recognition using human skeleton provided by Kinect," 2013 International Conference on Computing, Management and Telecommunications (ComManTel), Ho Chi Minh City, Vietnam, 2013, pp. 340-345.
[17]X. Chen, Q. Huang, W. Zhang, Z. Yu, R. Li and P. Lv, "Ping-pong trajectory perception and prediction by a PC based High speed four-camera vision system," 2011 9th World Congress on Intelligent Control and Automation, Taipei, 2011, pp. 1087-1092.
[18]Yong Qin, "Analysis of rotating dynamics of table tennis," 2011 International Conference on Computer Science and Service System (CSSS), Nanjing, 2011, pp. 3038-3041.