臺灣博碩士論文加值系統

本論文探討一個適用於體育課程的動作分類教練系統。傳統的運動課程需要教練進行一對一的教學，在一個大型班級對每個學生進行一對一的教學需要相當多的時間和精力。目前電腦輔助教學 (CAI) 在數位化教學的潮流下已經越來越普遍了。但是電腦輔助教學通常被限制在文字形式的教授，譬如數學、語言課程。因此發展一個適用於運動課的動作訓練系統是非常必需的。本文使用Wii控制器 (Wiimote)建構了一個低成本的動作擷取系統，本系統適用於網球、桌球等運動的基礎動作教練。本系統使用Wiimote感測肢體的加速度，如此每個Wiimote可以藉由藍芽傳輸傳回每個肢體的加速度資訊。在擷取完肢體各個部位的加速度資訊完後，電腦使用這些資訊辨識動作並將其分類到正確和錯誤的動作型態。因而這個系統可以提供適當的建議給使用者。這個系統使用修改過的連續數值ID3歸納學習法去產生決策樹。本論文發展出一個直覺的視覺化界面給教練以及學生使用。實驗結果指出平均的辨識動作錯誤的準確率為83%。這個系統減輕教練的工作負擔以及改進學生的學習成效。

This thesis proposed and analyzed a movement training system aiming to classify motions for physical education is proposed and analyzed. Traditional physical education requires an instructor teaching exercise movement individually. Teaching every student in a big class demands considerable time and efforts. Computer assisted instruction (CAI) is able to decrease workload. However, CAI is often confined in literal course such as mathematics, language courses. It is necessary to develop a motion training system for physical education. This thesis develops a low-cost motion capture with Wii Remote Control (Wiimote) for training movement exercise, such as tennis and baseball. Each Wiimote is attached to the limb, and then send back the acceleration information to the computer via Bluetooth wireless link. Thus the computer recognizes the motion and classifies the motion to several correct and incorrect categories. As a result, it is able to provide the appropriate advice to the students. The system applies a modified ID3 inductive learning to generate a decision tree with continuous-valued attributes. This thesis develops an easy-to-use GUI interface for coaches. The results show that the average accuracy of classification is 83%. The system reduces the workload of the coach and improves teaching and learning performance.

中文摘要 i
Abstract ii
誌謝 iii
Contents iv
List of Figures vii
List of Tables viii
Chapter 1 Introduction 1
1.1 Background and motivations 1
1.2 Objectives 3
1.3 Thesis organization 4
Chapter 2 Related Works 6
2.1 Decision tree 6
2.2 Motion capture 7
2.3 Motion classification and training 10
2.4 Wiimote control 12
2.5 Technical analysis on movement 13
Chapter 3 System Architecture 16
3.1 System overview 16
3.2 Motion capture subsystem 19
3.3 Decision tree model 21
Chapter 4 Research Methodology 24
4.1 Categories of forehand swing movement 24
4.2 Experiment equipments 27
4.3 The preparations of the experiment 29
4.4 The process of the training 29
Chapter 5 Experiment Results 31
5.1 Accelerations in the swing motions of each categories 31
5.2 Classification of each movement 38
5.3 Accuracy of the classifier 39
Chapter 6 Conclusions and Future Works 41
6.1 Conclusions 41
6.2 Future works 42
References 43
Appendix A: Source Code of ID3 45

[1]. A Shashua, Y Gdalyahu, G Hayun “Pedestrian Detection for Driving Assistance Systems: Single-frame Classification and System Level Performance”, Proceedings of the 2004 IEEE Intelligent Vehicles Symposium, 2004
[2]. D. Kwon, M. Gross, “Combining Body Sensors and Visual Sensors for Motion Training”, Proceedings of the 2005 ACM SIGCHI International Conference on Advances in computer entertainment technology, Valencia pp. 94-101 2005
[3]. D. Vlasic, R. Adelsberger, G Vannucci, J Barnwell et al, “Practical Motion Capture in Everyday Surroundings”, ACM Transactions on Graphics (TOG), Volume 26 , Issue 3, 2007
[4]. F. Multon, L. Hoyet, T. Komura and R. Kulpa, “Interactive control of physically-valid aerial motion: application to VR training system for gymnasts”, Proceedings of 2007 ACM symposium on Virtual reality software and technology, USA, pp. 77-80, 2007.
[5]. G. Liu, J Zhang, W. Wang, L. McMillan, “A System for Analyzing and Indexing Human-Motion Databases”, Proceedings of the 2005 ACM SIGMOD, pp. 924-926,2005.
[6]. H. Liu, S Lin, Y Zhang and K Tao, “Automatic Video-based Analysis of Athlete Action”, Proceedings of Image Analysis and Processing, 2007. ICIAP 2007. 14th International Conference, Modena, pp. 205-210, September 2007.
[7]. J. Foody, D. Kelly, D Kumar, D Fitzgerald et al., “A Real Time Motion Capture System, using USB Based Tri-Axis Magnetic and Inertial Sensors, for Movement Based Relaxation”, Proceedings of Irish Signals and Systems Conference, pp. 139-142, 2006
[8]. K. Hachimura, H. Kato and H Tamura, “A Prototype Dance Training Support System with Motion Capture and Mixed Reality Technologies”, Proceedings of 13th IEEE International Workshop on Robot and Human, Roman, pp. 217-222, September 2004.
[9]. L. Bao, S. Intille, “Activity Recognition from User-Annotated Acceleration Data”, Lecture Notes in Computer Science, 2004
[10]. Managed Library for Nintendo's Wiimote, http://blogs.msdn.com/coding4fun/archive /2007/03/14/1879033.aspx
[11]. P. C. Huang, “Finding multi-interval classification rules on continuous-valued attributes, a decision tree approach”, Institute of Information Management National Cheng Kung University. volume 3, 2003

[12]. R Slyper, J Hodgins, “Action Capture with Accelerometers”, Proceedings of SIGGRAPH Eurographics Symposium on Computer Animation, pp. 193-200, 2008
[13]. R. López, “A distance-based attribute selection measure for decision tree induction”, Machine Learning, Volume 6, Number 1, pp. 81-92, 1991.
[14]. S. Baek, S. Lee, and G. Jounghyun Kim “Motion Evaluation for VR-based Motion Training”, EUROGRAPHICS 2001. volume 3, 2002
[15]. T. Schlömer, B. Poppinga, N Henze and S Boll, “Gesture Recognition with a Wii Controller”, Proceedings of the 2nd international conference on Tangible and embedded interaction, Bonn, Germany, pp. 11-14, 2008.
[16]. V. B. Zordan and J. K. Hodgins, “Motion capture-driven simulations that hit and react”, Proceedings of the 2002 ACM SIGGRAPH/Eurographics symposium on Computer animation, Texas, pp. 89-96, 2002.
[17]. W. Zhou and A. Vellaikal, CCJ. Kuo, “Rule-based video classification system for basketball video indexing”, Proceedings of the 2000 ACM workshops on Multimedia, Bonn, Germany, pp. 213-216, 2000.