臺灣博碩士論文加值系統

本論文研究目的為將 CIELAB 色彩空間衍生之分色技術應用於 3D 掃描至3D 印表機過程，以建立 3D 影像之喜好或忠實的膚色再現處理，其中包含利用CIELAB 分色技術於3D 臉部模型之上色處理與彩妝模擬演算，與 3D 彩色列印系統之膚色修正方法。為了達成 3D 影像喜好膚色再現之目的，本論文共設計有三個實驗，實驗一為應用 CIELAB 分色技術提出的膚色移植法，解決了黑白臉部影像缺少膚色色彩資訊的問題，將成人參考影像之膚色移植至 2D 黑白孩童原始影像，再將原始2D 影像轉換至3D 影像。實驗二為整合並優化過去實驗室團隊所開發之彩妝模擬，解決無法全臉彩妝模擬的問題，將彩妝參考影像之彩妝顏色，優化混色並移植至 3D 原始影像，並整合使得3D 彩妝模擬系統更加完整；實驗二使用多項式回歸，解決了全彩 3D 印表機無法對膚色準確重現的問題，將平均色差E*ab 由 5.4 降至 4.0，最後將色彩修正處理後的 3D 臉部模型，列印輸出成真實的全彩 3D 臉部模型。

The aim of this study is to build a complete 3D coloring process with the skin-color preferred or skin-color fidelity intents that covers 3D scanning to 3D full-color printing. It can apply to the 3D scanning facial images by using CIELAB color separation technology, and 3D printing facial images by using color correction method. To achieve preferred skin-color reproduction of 3D images, three kinds of experiments were designed in this study. The first experiment proposed the skin-color grafting method. It can apply to the grayscale child facial-images in CIELAB color space and morph the 2D grayscale facial-images into the 3D coloring facial-images. Then, the second experiment solved the problems of the forehead without makeup in our first-generation 3D cosmetic simulation algorithm by means of the improved facial feature-point extraction. The third experiment established a color correction method for the full-color 3D printer system. By introducing the polynomial regression applying to 3D printer color correction. The total color differences were reduced and the average E*ab values before and after color corrections were 5.4 and 4.0 respectively. Finally, the color-corrected 3D facial images can print out as the 3D facial models with true colors by the output of the full-color 3D printer.

目次
摘要 II
ABSTRACT III
致謝 IV
目次 V
表目錄 VII
圖目錄 VIII
第一章 緒論 1
第二章 文獻回顧 3
2.1 人臉偵測 3
(1) 以人臉的特徵為基礎 3
(2) 模板匹配 4
(3) 類神經網路的方式 4
2.2 彩妝模擬系統 5
(1) 2D數位彩妝模擬 5
(2) 2D彩妝模擬系統 6
(3) 2D彩妝模擬系統 7
(4) 3D彩妝模擬系統 9
2.3 3D列印色彩系統 10
第三章 實驗原理 13
3.1 3D成像系統 13
(1) 3D掃描 13
(2) 3D影像格式 13
(3) 3D列印 14
3.2 演算法介紹 15
(1) 3D影像平面化 15
(2) 臉部特徵點處理 16
(3) CIELAB色彩空間轉換 18
(4) 多項式回歸 22
第四章 3D臉部上色演算法 24
4.1 實驗設備 25
4.2 實驗1 - 3D黑白影像臉部上色重建 26
(1) 3D臉部變形(映射) 26
(2) 膚色移植法 27
4.3 實驗2 – 3D彩妝模擬法 29
(1) 3D臉部演算法 30
(2) 彩妝模擬演算法 30
4.4 實驗3 - 3D印表機膚色修正 32
(1) 實驗流程 32
(2) 製作彩色導表與色域分析 33
(3) 製作膚色導表與比對色彩 37
(4) 色彩修正 37
(5) 彩妝應用 40
4.5 實驗結果 40
第五章 結論 44
參考文獻 45
附錄 膚色色彩修正表 48

參考文獻
[1] Craw, Ian, H. Ellis, and J. Rowland Lishman. "Automatic extraction of face-features." Pattern recognition letters 5.2 (1987): 183-187.
[2] Heisele, Bernd, et al. "Face recognition: component-based versus global approaches." Computer vision and image understanding 91.1 (2003): 6-21.
[3] Miao, Jun, et al. "A hierarchical multiscale and multiangle system for human face detection in a complex background using gravity-center template." Pattern Recognition 32.7 (1999): 1237-1248.
[4] Chuang, Jen-Hui. "A potential-based approach for shape matching and recognition." Pattern Recognition 29.3 (1996): 463-470.
[5] R. Brunelli and T. Poggio, “Face recognition: Features versus templates,” IEEE Trans. Pattern Analysis and Machine Intelligence, 1993, pp.1042-1052.
[6] Rowley, Henry A., Shumeet Baluja, and Takeo Kanade. "Rotation invariant neural network-based face detection." Computer Vision and Pattern Recognition, 1998. Proceedings. 1998 IEEE Computer Society Conference on. IEEE, 1998.
[7] Huang, Lin-Lin, et al. "Face detection from cluttered images using a polynomial neural network." Neurocomputing 51 (2003): 197-211.
[8] Dalal, Navneet, and Bill Triggs. "Histograms of oriented gradients for human detection." Computer Vision and Pattern Recognition, 2005. CVPR 2005. IEEE Computer Society Conference on. Vol. 1. IEEE, 2005.
[9] King, D. E. (2009). Dlib-ml: A machine learning toolkit. Journal of Machine Learning Research, 10(Jul), 1755-1758
[10] Tong, Wai-Shun, et al. "Example-based cosmetic transfer." Computer Graphics and Applications, 2007. PG'07. 15th Pacific Conference on. IEEE, 2007.
[11] Guo, Dong, and Terence Sim. "Digital face makeup by example." Computer Vision and Pattern Recognition, 2009. CVPR 2009. IEEE Conference on. IEEE, 2009.
[12]盧霖,「開發臉部影像特徵點為中心之膚色對應演算法」,碩士論文,國立台灣科技大學,台北(2015)
[13] 皮遠韸,「3D臉部彩妝模擬技術」,碩士論文,國立台灣科技大學,台北(2016)
[14] Stanić, M., and B. Lozo. "Color and permanence issues in 3D ink-jet printing." MIPRO, 2010 Proceedings of the 33rd International Convention. IEEE, 2010.
[15] Stanic, Maja, and Branka Lozo. "Permanence and Color Stability in 3D Ink-jet Printing." NIP & Digital Fabrication Conference. Vol. 2010. No. 2. Society for Imaging Science and Technology, 2010.
[16] Brunton, Alan, Can Ates Arikan, and Philipp Urban. "Pushing the limits of 3d color printing: Error diffusion with translucent materials." ACM Transactions on Graphics (TOG) 35.1 (2015): 4.
[17] Xiao, Kaida, et al. "Developing a 3D colour image reproduction system for additive manufacturing of facial prostheses." The International Journal of Advanced Manufacturing Technology 70.9-12 (2014): 2043-2049.
[18] Murray, J., and W. Van Ryper. "Wavefront OBJ File Format Summary." (2005).
[19] Bourke, P. (2009). Ply-polygon file format. Dostupné z http://paulbourke. net/dataformats/ply. (2017)
[20]吳懷宇編著,「決戰3D列印：「智造」時代來臨，顛覆你的既定「印」象」, 有意思出版社, 2015
[21] Beier, Thaddeus, and Shawn Neely. "Feature-based image metamorphosis." ACM SIGGRAPH Computer Graphics. Vol. 26. No. 2. ACM, 1992.
[22] 黃日鋒、詹文鑫、陳鴻興、胡國瑞、徐道義、孫沛立、羅梅君 編著,陳鴻興 編審, 「顯示色彩工程學」,全華圖書股份有限公司, 2011, 第139頁。
[23] Ohta, Noboru, and Alan Robertson. Colorimetry: fundamentals and applications. John Wiley & Sons, 2006.
[24] Montgomery, Douglas C., Elizabeth A. Peck, and G. Geoffrey Vining. Introduction to linear regression analysis. John Wiley & Sons, 2015.
[25] Hong, Guowei, M. Ronnier Luo, and Peter A. Rhodes. "A study of digital camera colorimetric characterisation based on polynomial modelling." (2001).
[26] Han, Kyuseo, et al. "Characterizing the desktop color printer with polynomial regression." Systems, Man, and Cybernetics, 1998. 1998 IEEE International Conference on. Vol. 5. IEEE, 1998.