近年來，人機互動議題愈來愈受到重視，眾多相關研究不斷地被發表出來，成為當代顯學；而在當中，手勢互動為最熱門的項目之一，用手勢控制來取代傳統的鍵盤滑鼠更是一種趨勢。本論文將實作一套手勢辨識系統，使用Leap Motion Controller作為感測器，擷取自定的手部特徵並透過機器學習Multinomial Logistic Regression演算法進行學習與辨識，辨認出10種手勢。
透過本論文所實作的方法，整體之正確辨識率可達98%；此外，由於Multinomial Logistic Regression所產生的預測模型具有著運算簡單之數學特性，因此本系統所耗費的資源低，不但可以表現出即時辨識的效能，更有著未來移植到嵌入式系統的可能性。另外，本系統未來也可應用在虛擬鍵盤滑鼠、手部復健等用途上，讓使用者有著更佳的使用體驗。

In recent years, more and more attention has been paid to Human-Computer Interaction issues, many related studies have been published. Among the issues, Gesture Interaction is one of the most popular studies; it is almost become a trend to use gesture control technique to replace the keyboard and mouse. This thesis proposes a hand gesture recognition system, using Leap Motion Controller as a sensor, capture the features of hands and calculate the data through the Multinomial Logistic Regression algorithm in order to get the Prediction Model to classify gestures into ten kinds of gestures.
The method we propose has average recognition rate of 98%. Moreover, with the benefit of low complexity of the machine learning method we use, our system not only has the real-time performance but also is possible to run in the embedded systems. In addition, our system can also be used for the purpose of virtual keyboard or mouse, hand rehabilitation and other way to make users have a better experience.

論文審定書 i
誌謝 ii
摘要 iii
Abstract iv
目錄 v
圖次 vii
表次 viii
第一章 序論 1
1.1研究動機與目的 1
1.2研究議題描述 2
1.3論文架構 5
第二章 背景知識 6
2.1 手勢種類介紹 6
2.1.1 靜態手勢 6
2.1.2 動態手勢 7
2.2 相關研究 8
2.3 Regression 9
2.3.1 線性迴歸 9
2.3.2 Logistic Regression 10
2.3.3 Multinomial Logistic Regression 13
2.4 支持向量機（SVM） 18
第三章 Leap Motion 21
3.1 Leap Motion Controller 21
3.2 Leap Motion SDK 23
3.3 Leap Motion Visualizer 26
第四章 研究方法 27
4.1 系統架構 27
4.2 系統流程 28
4.2.1 Training流程 28
4.2.2 Recognition流程 29
4.3 系統環境 30
4.3.1 開發環境 30
4.3.2 測試環境 30
第五章 系統實作 31
5.1 特徵擷取 32
5.2 Training Process 35
5.3 Recognition Process 38
5.3.1 機率運算 38
5.3.2 取樣方法 41
第六章 實驗數據與系統成果分析 42
6.1實驗方法 42
6.2 Multinomial Logistic Regression實驗結果 42
6.3 SVM實驗結果 43
6.4 兩方法比較 44
第七章 結論與未來展望 46
參考文獻 48
附錄一：學位考試委員問題與回覆 53
附錄二：投稿論文 55

[1] Alan Dix, Janet Finlay, Gregory D. Abowd, and Russell Beale, Human-Computer Interaction. Available: https://www.researchgate.net/profile/Alan_Dix/publication/224927543_Human-Computer_Interaction/links/02e7e51a84759ab04d000000/Human-Computer-Interaction.pdf, (2016-08-14)
[2] Leap Motion. Available: https://www.leapmotion.com/, (2016-03-11)
[3] Sérgio Bessa Carneiro, Edson D. F. De M. Santos, Talles M. De A. Barbosa, José O. Ferreira, and Adson F. Da Rocha, "Static Gestures Recognition for Brazilian Sign Language with Kinect and Eigenhands", in International Journal of Computer Applications, 2016, pp. 35 – 38.
[4] G. Marin, F. Dominio, and P. Zanuttigh, "Hand gesture recognition with leap motion and kinect devices", in IEEE International Conference on Image Processing (ICIP), 2014, pp. 1565 – 1569.
[5] G. Plouffe, and A. M. Cretu, "Static and Dynamic Hand Gesture Recognition in Depth Data Using Dynamic Time Warping", in IEEE Transactions on Instrumentation and Measurement, 2016, pp. 305 – 316.
[6] Hamid A. Jalab, "Static Hand Gesture Recognition for Human Computer Interaction", in Information Technology Journal, 2012, pp. 1265 – 1271.
[7] Y. Kim, and B. Toomajian, "Application of Doppler radar for the recognition of hand gestures using optimized deep convolutional neural networks", in 2017 11th European Conference on Antennas and Propagation (EUCAP), 2017, pp. 1258 – 1260.
[8] D. K. Ghosh, and S. Ari, "Static Hand Gesture Recognition Using Mixture of Features and SVM Classifier", in 2015 Fifth International Conference on Communication Systems and Network Technologies, 2015, pp. 1094 – 1099.
[9] M. K. Sohn, S. H. Lee, H. Kim, and H. Park, "Enhanced hand part classification from a single depth image using random decision forests", in IET Computer Vision, 2016, pp. 861 – 867.
[10] DTREG, A Brief History of Neural Networks. Available: https://www.dtreg.com/solution/view/21, (2017-07-12)
[11] Random forests - classification description. Available: https://www.stat.berkeley.edu/~breiman/RandomForests/cc_home.htm, (2017-07-29)
[12] Devendrakumar H. Pal, and S. M. Kakade, "Dynamic hand gesture recognition using kinect sensor", in 2016 International Conference on Global Trends in Signal Processing, Information Computing and Communication (ICGTSPICC), 2016, pp. 448 – 453.
[13] Sean R Eddy, "Profile hidden Markov models", in Bioinformatics (Oxford, England), 1998, pp. 755 – 763.
[14] 許明翔, "具有深度學習精神之人類手勢影像辨識系統," 碩士, 自動化科技研究所, 國立臺北科技大學, 台北市, 2015.
[15] 邏輯回歸Logistic Regression. Available: http://140.128.36.81/2013cel/逢甲商圈問卷調查/邏輯斯迴歸（Logistic Regression）.htm, (2017-03-14)
[16] Dayton, C. M, "Logistic regression analysis", in Stat, 1992, pp. 474 – 574.
[17] T. Cleophas, and A. Zwinderman, (2013). Machine learning in medicine. Dordrecht [u.a.]: Springer, pp. 17 – 24.
[18] Wikipedia, Logistic regression. Available: https://en.wikipedia.org/wiki/Logistic_regression, (2017-03-31)
[19] 解讀 logistic regression, Available: https://researcher20.com/2010/06/02/解讀-logistic-regression/, (2017-03-20)
[20] V. Bobic, P. Tadic, and G. Kvascev, "Hand gesture recognition using neural network based techniques", in 2016 13th Symposium on Neural Networks and Applications (NEUREL), 2016, pp. 1 – 4.
[21] Wikipedia, Multinomial logistic regression. Available: https://en.wikipedia.org/wiki/Multinomial_logistic_regression, (2017-03-19)
[22] H. Cheng, L. Yang, and Z. Liu, "A Survey on 3D Hand Gesture Recognition", in IEEE Transactions on Circuits and Systems for Video Technology, 2016, pp. 1659 – 1673.
[23] UFLDL, Softmax regression, Available: http://deeplearning.stanford.edu/wiki/index.php/Softmax_Regression, (2017-03-22)
[24] STAN, Available: http://mc-stan.org/, (2016-11-27)
[25] 林宗勳, Support Vector Machines簡介. Available: http://www.cmlab.csie.ntu.edu.tw/~cyy/learning/tutorials/SVM2.pdf, (2017-07-01)
[26] Analytics Vidhya, Understanding Support Vector Machine algorithm from examples (along with code), Available: https://www.analyticsvidhya.com/blog/2015/10/understaing-support-vector-machine-example-code/, (2017-07-10)
[27] LIBSVM -- A Library for Support Vector Machines, Available: https://www.csie.ntu.edu.tw/~cjlin/libsvm/, (2017-05-22)
[28] Leap Motion, Leap Motion C++ SDK V2.0 Beta documentation, Available: http://users.itk.ppke.hu/~horan/LeapImageJ/LeapSDK/docs/cpp/api/Leap.Device.html, (2017-07-05)
[29] Leap Motion Blog, Available: http://blog.leapmotion.com/getting-started-leap-motion-sdk/, (2017-07-05)
[30] Y. Chen, Z. Ding, Y. L. Chen, and X. Wu, "Rapid recognition of dynamic hand gestures using leap motion", in 2015 IEEE International Conference on Information and Automation, 2015, pp. 1419 – 1424.
[31] W. Lu, Z. Tong, and J. Chu, "Dynamic Hand Gesture Recognition With Leap Motion Controller", in IEEE Signal Processing Letters, 2016, pp. 1188 – 1192.
[32] J. K. Sharma, R. Gupta, and V. K. Pathak, "Numeral Gesture Recognition using Leap Motion Sensor", in 2015 International Conference on Computational Intelligence and Communication Networks (CICN), 2015, pp. 411 – 414.
[33] Tensorflow, MNIST For ML Beginners, Available: https://www.tensorflow.org/get_started/mnist/beginners, (2017-07-10)
[34] Kdnuggets, What is Softmax Regression and How is it Related to Logistic Regression?, Available: http://www.kdnuggets.com/2016/07/softmax-regression-related-logistic-regression.html, (2017-03-20)
[35] J. Gałka, M. Mąsior, M. Zaborski, and K. Barczewska, "Inertial Motion Sensing Glove for Sign Language Gesture Acquisition and Recognition", in IEEE Sensors Journal, 2016, pp. 6310 – 6316.
[36] S. Lee, K. Park, J. Lee, and K. Kim, "User Study of VR Basic Controller and Data Glove as Hand Gesture Inputs in VR Games", in 2017 International Symposium on Ubiquitous Virtual Reality (ISUVR), 2017, pp. 1 – 3.
[37] S. Kudubayeva, D. Ryumin, M. Kalghanov, and A. Assem, "Automated Recognition System of Statistic Gestures via Kinect Sensor", in 2016 IEEE 10th International Conference on Application of Information and Communication Technologies (AICT), 2016, pp. 1 – 4.
[38] C. Miron, A. Păsărică, D. Arotăriţei, H. Costin, R. G. Bozomitu, and C. Rotariu, "Hand gesture detection using a stereo camera system and simulation of movement", in 2017 10th International Symposium on Advanced Topics in Electrical Engineering (ATEE), 2017, pp. 297 – 300.