臺灣博碩士論文加值系統

利用2D影像重建3D人臉一直是受到重視的研究領域之ㄧ。本論文有鑒於3D人臉重建之相關應用日益增多，從最熱門電影科技產業到走向具有藝術收藏價值的公仔、3D水晶飾品，發展出一套快速之3D人臉重建系統，以兩張2D影像配合一個基礎人模為條件，經過影像與人模前處理後，以Morphing為核心演算法藉由影像之特徵位置來變形人模之外觀。最後針對不足之處，後續由自由造型變形弁鄖荍硈t調整差異，最後重建出個性化之3D人臉。
2D影像之特殊影像處理將人臉影像轉換為以點密度疏密分佈呈現之人臉，配合一事先準備的標準人模，以Morphing快速產生目的人臉，不同於一般人臉研究最後以貼圖技術呈現結果，本研究最後所呈現的為ㄧ個3D空間點資料分佈之人臉，並以Bezier與B-Spline數學模型為主建構出具有整體變形與局部變形之編修工具，使得重建出之3D人臉有更豐富的變化。
歸納上述所言，本系統除了節省硬體成本之外，透過編修弁鄍i以增加產品之多樣性、提昇競爭能力。

Technology capable of generating a realistic three-dimensional human face model using 2D images and a generic model has received a lot of attention. Because of the growth of 3D face application software, including for animation and artistic production, we have also developed a system that can reconstruct a 3D human face model easily. The system takes a pair of 2D face images and a generic 3D face model with corresponding feature points. The Morphing Algorithm is then used to reconstruct the 3D face. The system includes an editing tool that deforms the resulting 3D face model using the Free-Form Deformation (FFD) technique.
Our proposed system is different from other research on 3D face modeling in that our imaging process converts a 2D personal image to a 3D point-based data that displays a 3D face with multi-density point clouds. The FFD editing tool with Bezier and B-Spline mathematic modelling are used to deform the resulting 3D face through global and local editing.
In conclusion, our system not only reconstructs a 3D face model easily but also saves on hardware cost. The editing tool enables the results to have more variety and be more interesting.

目錄
中文摘要..................................................................................................Ⅰ
英文摘要..................................................................................................Ⅱ
目錄........................................................................................................III圖目錄..................................................................................................... VI

第一章 緒論.................................................................................... 001
1.1 前言...................................................................................... 001
1.2 研究動機.............................................................................. 001
1.3 研究目的….......................................................................... 003
1.4 文獻回顧.............................................................................. 004
1.4.1 重建3D人臉方式總覽................................................ 004
1.4.2 影像重建3D人臉分類................................................ 006
1.4.3 以特徵點方式重建3D人臉........................................ 010
1.4.4 虛擬人臉表情模擬與複製...........................................011
1.5 3D人臉重建應用探討...........................................................012
1.6 研究方法與章節簡介............................................................014
第二章 影像前處理.....................................................................017
2.1 影像亮度/對比調配、背景處理.............................................017
2.2 影像銳利化處理....................................................................018
2.3 影像灰階轉換與佈點處理....................................................021
第三章 D基礎人模前處理.....................................................022
3.1 Surfel 介紹.............................................................................023
3.2 亂點資料之鄰近點搜尋........................................................024
3.2.1 固定範圍搜尋鄰近點關係..........................................024
3.2.2 以最近點個數法尋找鄰近點關係..............................025
3.3.3 距離法與最近點個數法之結果比較......................... 026
3.3 最小平方法擬合法向量......................................................027
3.4 Surfel半徑計算.....................................................................029
第四章 變形演算法基礎理論介紹.........................................031
4.1 扭曲(Warping)的介紹...........................................................032
4.2 扭曲(Warping)的演算法.......................................................034
4.3 G Function的選擇.................................................................037
4.4 初定位...................................................................................039
第五章 側邊輪廓自動偵測與擬合.........................................042
5.1 人臉側邊特徵分析與條件限制...........................................042
5.2 基礎人模輪廓判定與特徵偵測...........................................047
5.3 2D影像側邊輪廓偵測結果.................................................048
5.4 側邊擬合...............................................................................050
第六章 3D人臉編修...................................................................052
6.1 Free-Form Defromation 介紹................................................052
6.2 Free-Form Deformation 原理................................................053
6.2.1 Bezier模型介紹............................................................053
6.2.2 B-spline 模型介紹.......................................................058
6.3 FFD 應用於人臉之編修…………………..…………….…..063
第七章 3D人臉重建流程..........................................................063
7.1 2D影像處理流程..................................................................066
7.2 3D基礎人模處理流程..........................................................067
7.3 手動特徵萃取.......................................................................067
7.4 單一2D影像重建3D人臉...................................................071
7.5 正面與側面影像手動重建3D人臉.....................................073
7.6 正面與側面影像半自動重建3D人臉.................................075
第八章 結論與未來工作............................................................078
8.1 結論.......................................................................................078
8.2 未來方向...............................................................................079
參考文獻..............................................................................................081

參考文獻：

[1] W. S. Lee, N. M. Thalmann, “Fast head modelling for animation”
, Journal Image and Vision Computing, vol. 18, no. 4, pp. 355-364
, 2000.
[2]B. Moghaddam, J. Lee, H. Pfister, and R. Machiraju, “Model-based 3-D face capture with shape-from-silhouettes”, In IEEE International Workshop on Analysis and Modeling of Faces and Gestures, Nice, France, pp. 20–27, 2003.
[3]N. Rasiwasia, and K. S. Venkatesh, “The Avatar: 3-D Face reconstruction from two orthogonal pictureswith application to facial makeover”.
[4]A. N. Ansari, and A. M. Mohamed, “3-D Face Modeling Using Two Views and a Generic Face Model with Application to 3-D Face Recognition”, In IEEE Conference on Advanced Video and Signal Based Surveillance (AVSS'03), 2003.
[5]S. Shan, W. Gao, J. Yan, H. Zhang, and X. Chen, “Individual 3D face synthesis based on orthogonal photos and speech-driven facial animation”, In Proc. Int. Conf. Image Process (ICIP2000),
pp. 238–241, 2000.
[6]F. Pighin, J. Hecker, D. Lischinski, Szeliski R, and D. Salesin. “Synthesizing realistic facial expressions from photographs”, In Computer Graphics Proceedings SIGGRAPH’98, pp. 75–84, 1998.
[7] M. Castel’an, and E. R. Hancock, “Acquiring height maps of
faces from a single image”, In Proc. 3DPVT, pp. 183–190,
2004.
[8] J. Atick, P. Griffin, and A. Redlich, “Statistical approach to shape from shading: Reconstruction of 3D face surfaces from single 2D images”, Neural Computation, 1996.
[9] L. Lee, B. Moghaddam, H. P_ster, and R. Machiraju,
“Silhouette-based 3D face shape recovery”, Graphics Interface,
2003.
[10] V. Blanz and T. Vetter, “Face Recognition Based on Fitting a
3D Morphable Model”, IEEE Transactions on Pattern Analysis
and Machine Intelligence, vol. 25, no. 3, 2003.


[11] N. Sams, N. Grammalidis, and M.G.Strintzis, “Building
three-dimensional head models”, Graphical Models,
vol. 63, no. 5, pp. 333-368, 2001.
[12] A. R. Chowdhury, R. Chellappa, R. Krishnamurthy and
T. Vo, “3D Face Recostruction from Video Using A Generic
Model”, In Proc. of Int. Conf. on Multimedia and Expo,
Lausanne, Switzerland, 2002.
[13] J. Y. Noh and U. Neumann, ”Expression Cloning”, In Proc. of ACM SIGGRAPH’01, pp. 277-288, 2001.
[14] H. Pyun, Y. Kim, W. Chae, H. W. Kang, and S. Y. Shin, “An example-based approach for facial expression cloning”,
In Proc. of SIGGRAPH/EG Symposium on Computer Animation (SCA), pp. 167-176,2003.

[15] R. C. Gnzalez, and R. E. Woods, “Digital Image Processing”, 2001.
[16] H. Pfister, M. Zwicker, J. van Baar, and M. Gross, ” Surfels: Surface Elements as Rendering Primitives”, In Computer Graphics, SIGGRAPH 2000 Proceedings, pp. 335–342, 2000.
[17]M. B. Kobbelt, “A Survey of Point-Based Techniques in Computer Graphics”, 2004.
[18] http://bit.kuas.edu.tw/~cscheng/research/paper/kdtree.htm
[19] Webster ,”Webster’s Encyclopedic Unabridged Dictionary of the English Language”, New York: Random House , 1989.
[20] http://www.stanford.edu/~jacobliu/368Report/ee368_finalReport
.html
[21] D. Levin, and A. SoloMouivi, “Three-Dimensional Distance Field Metamorphosis”,ACM Transactions on Graphics, vol. 17, no. 2, pp. 116-141, 1998.
[22] J. Noh, D. Fidaleo, and U. Neumann, ”Animated deformations
with radial basis functions”, ACM Symposium on Virtual Reality Software and Technology (VRST2000), pp. 166–174, 2000.
[23] 劉育呈, “自動化整修具限制條件的自由曲面模型”,中正機械所, 1999.


[24] H. Gupta, A. K. RoyChowdhury, R. Chellappa, “Contour-based 3D Face Modeling from a Monocular Video”, The British Machine Vision Conference, 2004.
[25] T. W. Sederberg, and S. R. Parry, “Free Form Deformations of Solid Geometric Models.”, Proceedings of Siggraph 86. In Computer Graphics vol. 20, no. 4 , 1986.
[26] F. Alex, “Real-Time Soft-Object Animation Using Free-Form Deformation.”, 2003.
[27]http://www.gdmag.com