5-1 結論
人類利用各式各樣的裝置控制機器，越來越多的輸入介面不斷發展，
而利用體感遙控正是人類夢寐以求的方式，我們創造了一套姿勢系統，
使用此姿勢辨識系統能夠準確地進行辨識。
利用動態扭曲時間演算法，搭配 Kinect 深度感應器的骨架追蹤，將 骨架關節資訊分別以 DTW 演算法運算，將參考資料與輸入資料比對，
以達到姿勢辨識的目標。
本論文第一次提出利用 DTW 動態扭曲時間演算法，搭配 Kinect 的 動態骨架追蹤建構出一套精確的姿勢辨識系統，不受距離以及動作速 度的影響，並且第一次針對辨識系統進行一連串分析，本系統姿勢辨 識準確度高達 70%以上，而較單純的動作，例如左手往左邊揮動，辨 識準確度甚至能達到 90%以上，若辨識者與紀錄者為同一人時，辨識 的成功率幾乎為 100%。
利用高辨識準確度的特性，在本系統上操作許多預設的功能，能夠 輕鬆的進行體感遙控。
- 43 -
5-2 未來展望
姿勢體感遙控的趨勢已經勢不可擋，利用 3D 的體感遙控與平面顯示 器進行遊戲互動開啟了輸入界面的新視野。
需要高度準確的姿勢辨識，本論文系統提供一個最佳的範例，而未 來體感更可突破與平面顯示器的互動，進一步利用體感控制實體物品，
例如機器手臂、遙控汽車等等，都將帶給使用者全新不同的感受。
有許多領域需要姿勢辨識，舉例來說:運動賽事裁判、領航員等等，
本辨識系統若能建立準確姿勢資料庫，相信能提供莫大的幫助，不但 能降低人員成本支出，更能減少人員判斷錯誤的風險。
- 44 -
參考文獻
[1] 劉美君等編著，3D影像互動應用趨勢與系統解決方案探索，經濟 部發行，民國100年。
[2] Thomas R. Moore，Ashley E. Cannaday, "Imaging of vibrating objects using speckle subtraction," Optical Society of America, Vol. 27, No. 8，
July28, 2010.
[3] Lee, J.S, "Speckle Analysis and Smoothing of Synthetic Aperture Radar Images," Computer Graphics and Image Processing, Vol.
17:24-32, 1981.
[4] M. Mansourpour, M.A. Rajabi , J.A.R. Blais , ”Effects and
performance of speckle noise reduction dilters on active radar and SAR images,” 7 January,2011.
[5] A. R. Fercher, “Velocity measurement by first order statistics of time-differentiated laser speckles,” Opt. Common,33, 129–135, 1980.
[6] Hitoshi Fujii, John W.Y. Lit, ”Surface Roughness Measurement Using Dichromatic Speckle Pattern: An Experimental Study,”APPLIED OPTICS ,Vol. 17, No. 17,1 September 1978.
[7] Bing Pan, Huimin Xie, Jianxin Gao, Anand Asundi, ”Improved speckle projection profilometry for out-of –plane shape measurement,”
APPLIED OPTICS,Vol.47, No.29, 10 October 2008.
[8] J García, Z Zalevsky, ” Projection of speckle patterns for 3D sensing,”
Journal of Physics, Conference Series 139, 2008.
[9] Abhishek Kar, ”Skeletal Tracking using Microsoft Kinect,”
[10] Valentino Frat, Domenico Prattichizzo, ”Using Kinect for hand tracking and rendering in wearable haptics,” IEEE World Haptics Conference 2011, 21-24 June, 2011.
[11] Shahram Izadi, Otmar Hilliges, ” KinectFusion: Real-time 3D Reconstruction and Interaction Using a Moving Depth Camera,”
Microsoft Research, 11 October 2011.
[12] Lawrence R.Rabiner , ”A tutorial on Hidden Markov Models and selected applications in speech recognition,” IEEE,Vol.77 ,No.2, 1989.
[13] 吳妍潔，「使用分層隱藏式馬可夫模型作人類動作辨識」，國立交
- 45 -
通大學，碩士論文，民國九十六年。
[14] Leon O.Chua, ”Cellular neural networks:Theory,” IEEE,Vol.35, No.10, October 1988.
[15] Guoqiang Peter Zhang, ”Neural networks for classification: A Survey,”
IEEE, Vol.30, No.4, November 2000.
[16] Hongjun Lu, ”Effective data mining using neural networks,”
IEEE,Vol.8, No.6, December 1996.
[17] Hiroaki Sakoe, Seibi Chiba, ”Dynamic Programming Algorithm Optimization for Spoken Word Recognition,” IEEE, Vol.ASSP-26, No.1, February 1978.
[18] Eamonn Keogh, Chotirat Ann Ratanamahatana, ”Exact indexing of dynamic time warping,” Knowledge and Information Systems ,2004.
[19] http://www.xbox.com/zh-TW/kinect
[20] Zeev Zalevsky, Alexander Shpunt, ”THREE-DIMENSIONAL SENSING USING SPECKLE PATTERNS,” US Patent, 2009.
[21] Stephen G. Latta, et al, ” GESTURE KEYBOARDING,” US Patent, 2010.
[22] Alexander Shpunt, Zeev Zalevsky, ” DEPTH-VARYING LIGHT FIELDS FOR THREE DIMENSIONAL SENSING,” US Patent, 2008.
[23] Javier Garcia, Zalevsky, ” RANGE MAPPING USING SPECKLE DECORRELATION,” US Patent , 2008.
[24] Roberta Piroddi, ” Stereoscopic 3D Technologies,” Snell Ltd, 2010.
[25] “The Business and Technology of Stereoscopic Filmmaking,”
Autodesk, 2010.
[26] MARIA TERESA BAEZA-ROMERO, et al, ” Time-of-Flight Mass Spectrometry for Time-Resolved Measurements: Some Developments and Applications,” DOI, 10.1002, 2011.
[27] Sergio Almansa-Valverde, et al, ” Mobile robot map building from time-of-flight camera,” Expert Systems with Applications , 39 8835–8843, 2012.
[28] Stefan May, et al, ”Three-Dimensional Mapping with Time-of-Flight Cameras,” Journal of Field Robotics , 26(11–12),934–965, 2009.
[29] Stefan Fuchs, Gerd Hirzinger, ”Extrinsic and Depth Calibration of ToF-cameras,” 2007.
[30] Michael Magee, et al, ”Location of features of known height in the presence of reflective and refractive noise using a stereoscopic light-striping approach,” SPIE, 10.1117/12.163139, 1994.
- 46 -
[31] Changsoo Je,et al, ” Colour-stripe permutation pattern for rapid
structured-light range imaging,” Optics Communications, 285 (2012) 2320–2331,2012.
[32] Jing Xu, et al,”Realtime 3D profile measurement by using the
composite pattern based on the binary stripe pattern,”Optics & Laser Technology ,44 (2012) 587–593,2012.
[33] Song Zhang,”Recent progresses on real-time 3D shape measurement using digital fringe projection techniques,” Optics and Lasers in Engineering, 48 (2010) 149–158,2010.
[34] U. Paul Kumar, et al,”Single frame digital fringe projection
profilometry for 3-D surface shape measurement,”Optik - Int. J. Light Electron Opt, doi:10.1016/j.ijleo.2011.11.030,2012.
[35] Alexander Shpunt, et al, ”Depth mapping based on pattern matching and stereoscopic information,” US Patent，2010.
[36] Feipeng Zhu,et al, ” Three-dimensional shape measurement and calibration for fringe projection by considering unequal height of the projector and the camera,” APPLIED OPTICS,10 April 2011 / Vol. 50, No. 11 ,2011.
[37] Chadi ALBITAR, ” Robust Structured Light Coding for 3D Reconstruction,” IEEE,978-1-4244-1631-8/07,2010.
[38] M.A. Sutton, et al,” The effect of out-of-plane motion on 2D and 3D digital image correlation measurements,” Optics and Lasers in
Engineering , 46 (2008) 746– 757,2008.
[39] Wei Sun,et al, ” 3D displacement measurement with a single camera based on digital image correlation technique,” Proc. of SPIE ,Vol.
6723 67230E-1,2007.
[40]
http://local.wasp.uwa.edu.au/~pbourke/miscellaneous/correlate/
[41] H.Lu, P.D.Cary, ”Deformation measurements by digital image
correlation :Implementation of a second-order displacement gradient,”
Experimental mechanics, 1999.
[42] C.J. Tay, et al,” Digital image correlation for whole field out-of-plane displacement measurement using a single camera,” Optics
Communications , 251 (2005) 23–36,2005.
[43] Bartosz Jablonski, ” Quaternion Dynamic Time Warping,” IEEE TRANSACTIONS ON SIGNAL PROCESSING,VOL. 60, NO.
3,2012.
[44] Lawrence R.Rabiner,et al,”Considerations in Dynamic Time
- 47 -
Warping Algorithm for Discrete Word Recognition,” IEEE,VOL.
ASSP-26, NO. 6,1978.
[45] Huaguang Zhang, et al, ”Data-Core-Based Fuzzy Min–Max Neural Network for Pattern Classification,” IEEE,VOL. 22, NO. 12,2011.
[46] Stelios D. Bekiros, ”Sign Prediction and Volatility Dynamics with Hybrid Neurofuzzy Approaches,” IEEE,VOL. 22, NO. 12,2011.
[47] Ya-Jun Qu,Bao-Gang Hu, ”Generalized Constraint Neural Network Regression Model Subject to Linear Priors,” IEEE,VOL. 22, NO.
12,2011.
[48] Jan Chorowski, Jacek M. Zurada,”Extracting Rules from Neural Networks as Decision Diagrams,” IEEE,VOL. 22, NO. 12,2011.
[49] Georgios Kotsalis, et al,” A Model Reduction Algorithm for Hidden Markov Models,” IEEE,1-4244-0171-2/06,2006.
[50] Wang Junli,et al,”Laser Speckle Images Research based on Wavelet-Domain Hidden Markov Models,” IEEE,978 -1-4244-9983-0/11, 2011.
[51] Quanwei Zhang,et al, ”Splice sites detection by combining Markov and hidden Markov model,” IEEE,978-1-4244-4134-1/09,2009.
H
.