本文提出一套自動年齡合成系統，可以自動估算輸入影像之年齡層，並執行臉型分類、五官對準及年齡皺紋之合成。自動化方式係以輸入影像經計算機預測之年齡，為合成影像的年齡基準，並整合ASM 演算法所得特徵點，為人臉影像校正及五官定位之基準，然後利用分類方法於資料庫中尋找最相似參考臉型後，以Log-Gabor小波抓取紋理特徵的方法，將人臉影像之年齡紋理解析出，從而合成出一系列不同年齡時期之影像。

This paper presents an automated synthesis system that can automati-cally estimate the age of the input image and perform facial classification, facial targeting and synthesis wrinkles of human ages. The computer auto-matically prediction the age of test image, that be the basis to synthetic im-ages, and will search the feature points of the face image by ASM algorithm, that will be the reference for facial component alignment. After that, we use the classifier to find the most similarly reference image in the database, and analysis the texture of the test image and reference image with Log-Gabor wavelet that can extraction the texture features about aging in-formation and wrinkle. Finally, we can simulate a series of facial images in the different period that will be younger or older than the test image.

目錄
致謝	I
中文摘要	III
英文摘要	IV
目錄	V
圖目錄	VII
第1章  緒論	1
1.1  研究背景	1
1.2  研究動機與目的	1
1.3  論文架構	2
第2章  文獻探討	4
2.1  基於非負矩陣分解演算法之人臉影像預測	4
2.1.1  非負矩陣分解演算法	4
2.1.2  運用於人臉影像預測之方法	6
2.2  基於多層解析動態模型之人臉影像預測	12
2.2.1  多層解析模型建立	12
2.2.2  動態模型建立	16
2.2.3  人臉老化模擬	18
2.3  基於Log-Gabor小波分析之人臉影像預測	23
2.3.1  AdaBoost演算法	23
2.3.2  ASM主動形狀模型	28
2.3.3  Log-Gabor小波分析	34
2.3.4  人臉老化影像預測	40
第3章  系統架構	44
3.1  影像前處理	46
3.1.1  特徵點萃取	46
3.1.2  旋轉校正	48
3.1.3  尺寸校正	49
3.2  臉型資料庫建立	51
3.2.1  年齡判斷	51
3.2.2  臉型分類	54
3.3  影像年齡合成	57
3.3.1  資料庫搜尋	57
3.3.2  人臉影像年齡模擬	58
第4章  實驗與討論	60
4.1  實驗環境	60
4.2  濾波器參數設定	60
4.3  實驗結果	67
第5章  結論與未來研究方向	71
参考文獻	72

圖目錄
圖2.1	許志維及張軒庭[7]應用NMF在人臉影像預測之方法	7
圖2.2	許志維及張軒庭[7]應用NMF訓練人臉基底影像之方法	9
圖2.3	許志維及張軒庭[7]應用NMF重建人臉影像之方法	10
圖2.4	許志維及張軒庭[7]應用NMF在人臉影像預測之模擬結果	11
圖2.5	Jinli Suo等人[10]提出之多層解析模型	14
圖2.6	Jinli Suo等人[10]提出之多層解析模型的部分樣本	15
圖2.7	Jinli Suo等人[10]提出之多層解析動態模型	16
圖2.8	Jinli Suo等人[10]提出多層解析動態模型之元件替換範例	20
圖2.9	Jinli Suo等人[10]提出多層解析動態模型之皺紋統計範例	21
圖2.10	Jinli Suo等人[10]提出多層解析動態模型之皺紋模擬結果	22
圖2.11	Jinli Suo等人[10]提出多層解析動態模型之人臉老化模擬結果	22
圖2.12	范聖恩[16]之Adaboost訓練演算法流程圖解	26
圖2.13	Paul Viola與Michael Jones[17]提出矩型特徵	27
圖2.14	Paul Viola與Michael Jones[17]提出之層疊分類原理	28
圖2.15	PDM控制點說明示意圖	29
圖2.16	主成份分析（PCA）示意圖	32
圖2.17	何景堂[21]運用Gabor filter萃取人臉紋理結果	36
圖2.18	頻率域中，Log-Gabor小波頻寬之形狀表示圖	39
圖2.19	空間域中，Log-Gabor小波頻寬形狀之實數表示圖	39
圖2.20	空間域中，Log-Gabor小波頻寬形狀之虛數表示圖	40
圖2.21	李祐承[13]及潘俊瑋[14]提出之人臉影像老化合成流程	41
圖2.22	李祐承[13]基於Log-Gabor小波提出之人臉影像老化結果	42
圖2.23	潘俊瑋[14]基於Log-Gabor小波提出之人臉影像老化結果	43
圖2.24	本文提出之系統方塊圖	45
圖3.1	ASM模型萃取75特徵點	47
圖3.2	特徵點描繪人臉輪廓及臉部元件之形狀及位置	47
圖3.3	內眼角不變性之旋轉校正	49
圖3.4	利用ASM特徵點之尺寸校正	50
圖3.5	改良李祐承[13]之年齡判斷過程	52
圖3.6	臉部元件位置關聯資訊之特徵點	52
圖3.7	臉部元件分類示意圖	52
圖3.8	資料庫搜尋示意圖	52
圖3.9	人臉影像年齡模擬方塊圖	52
圖4.1	不同中心頻率下，以Log-Gabor重建人臉影像	61
圖4.2	中心頻率（f0）及其8個方向（orientation,θ）之關係示意圖	63
圖4.3	中心頻率（f0）及其標準差（σf=0.05至0.30）之關係示意圖	64
圖4.4	中心頻率（f0）及其標準差（σf=0.35至0.60）之關係示意圖	65
圖4.5	尺度大小（scale）之紋理合成示意圖	66
圖4.6	年輕影像測試	68
圖4.7	中年影像測試	69
圖4.8	老年影像測試	70

[1]	S. Mukaida, H. Ando, “Extraction and Manipulation of Wrinkles and Spots for Facial Image Synthesis”, IEEE, International Conference, May 2004, pp. 749 – 754.
[2]	Jinli Suo, Song-Chun Zhu, Shiguang Shan, and Xilin Chen, “A Compositional and Dynamic Model for Face Aging.”, IEEE, Transactions on pattern analysis and machine intelligence, VOL. 32, NO. 3, March 2010.
[3]	Rowland, D.A., Perrett D.I., “Manipulating facial appearance through shape and color”, IEEE, Computer Graphics and Applications, Sept. 1995, pp. 70 – 76.
[4]	Lanitis, A., Taylor, C.J., Cootes and T.F., “Modeling the process of ageing in face images”, IEEE, International Conference on Computer Vision, Sept. 1999, pp. 131 – 136.
[5]	C. Choi, “Age change for predicting future faces”, IEEE, Fuzzy Systems Conference Proceedings, Aug. 1999, pp. 1603 – 1608.
[6]	Field D, “Relations between the Statistics of Natural Images and the Re-sponse Properties of Cortical Cells.”, Journal of Optical Society of American , 1987, pp. 2379-2393.
[7]	許志維,張軒庭,“基於非負矩陣分解演算法預測未來人臉影像”, Proceedings of the 2005 Workshop on Consumer Electronics and Signal Processing, 2005.
[8]	Daniel D. Lee and H. Sebastian Seung, “Learning the parts	of	objects	by	non-negative	matrix factorization.	Nature”,	1999, pp.788-791.
[9]	”Face and Gesture Recognition”, Network: FG-NET aging database, http://sting.cycollege.ac.cy/~alanitis/fgnetaging/
[10]	J. Suo, F. Min, S. C. Zhu, S. H. Shan, X. L. Chen, “A Multi-Resolution Dynamic Model for Face Aging Simulation”, IEEE, July, 2007.
[11]	Z. J. Xu, H. Chen, S. C. Zhu, “A High Resolution Grammatical Model for Face Representation and Sketching”, IEEE, Computer Society Conference on Computer Vision and Pattern Recognition, 2005.
[12]	T.F.Cootes, G.J. Edwards and C.J.Taylor. ”Active appearance models”, Proceedings of ECCV, 1998
[13]	李祐承, “Log-Gabor 小波為基礎之人臉合成系統”, 淡江大學, 電機工程研究所碩士論文, 2009.
[14]	潘俊瑋, “整合ASM與Log-Gabor小波之人臉合成系統”, 淡江大學, 電機工程研究所碩士論文, 2010.
[15]	Y. Freund, R. E. Schapire, “A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting,” Journal of Computer and System Sciences, 1997, pp. 119-139.
[16]	范聖恩, “以外形特徵為基礎之影像語言分類器-應用於破碎中文字合併”, 中央大學, 資訊工程研究所碩士論文, 2009.
[17]	P. Viola, M. Jones, “Robust Real Time Object Detection,” Second Interna-tional Workshop on Statistical and Computational Theories of Vision Vancouver, Canada, July 13, 2001.
[18]	Cootes, T.F., Taylor C, J., “Active shape models – smart snakes”, Proceedings British Machine Vision Conference, Springer, Berlin, 1992, pp. 266-275.
[19]	R.H. Davies, T.F. Cootes, C.Twining and C.J. Taylor” , An Information theoretic approach to statistical shape modelling”, Proc. British Machine Vision Conference, 2001, pp.3-11
[20]	Y. Pang, Y. Yuan, X. Li, “Gabor-Based Region Covariance Matrices for Face Recognition,” Circuits and Systems for Video Technology, IEEE Transactions, Vol. 18, July 2008, pp. 989 – 993.
[21]	何景堂, “以粒子群最佳化演算法為基礎之改良型多層類神經網路於臉部辨識應用”, 朝陽科技大學, 資訊工程研究所碩士論文, 2006。
[22]	S. M. Lajevardi, M. Lech, “Facial Expression Recognition Using Neural Networks and Log-Gabor Filters,” Computing: Techniques and Applications, 2008. DICTA '08.Digital Image, 1-3 Dec. 2008, pp. 77 – 83.
[23]	N. Rose, “Facial Expression Classification using Gabor and Log-Gabor Filters,” Automatic Face and Gesture Recognition, 2006. FGR 2006. 7th International Conference, 2-6 April 2006, pp. 346 - 350.
[24]	J. G. Daugman, “Uncertainty relations for resolution in space, spatial fre-quency, and orientation optimized by two-dimensional visual cortical filters,” Journal of the Optical Society of America, Vol. 2, 1985, pp. 1160-1169.
[25]	S. E. Grigorescu, N. Petkov, and P. Kruizinga, “Comparison of texture features based on Gabor filters,”  IEEE Trans. on Image Processing, Vol. 11, No. 10, 2002, pp. 1160-1167.
[26]	D. Dunn, W. E. Higgins, and J.  Wakeley, “Texture segmentation using 2-D Gabor elementary functions,” IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 16, No. 2, 1994.
[27]	S. Arivazhagan, L. Ganesan, and S. Bama, “Fault segmentation in fabric images using Gabor wavelet transform,” Machine Vision and Application, Vol. 16, No. 6, 2006, pp. 356-363.
[28]	A. Bodnarova, M. Bennamoun, and S. Latham, “Optimal Gabor filters for textile flaw detection,” Pattern Recognition, Vol. 35, No. 12, 2002, pp. 2973-2991.
[29]	D. M. Tsai, and S. K. Wu, “Automated surface inspection using gabor fil-ters,”  International Journal of  Advanced Manufacturing Technology, Vol. 16, No. 7, 2000, pp. 474-482.
[30]	鍾國亮，影像處理與電腦視覺，東華書局，2002。
[31]	R. C. Gonzalez, R. E. Woods, Digital Image Processing, 2nd ed., Pren-tice-Hall, 2002.
[32]	I. Rish, “An Empirical Study of The Naive Bayes Classifier”, Empirical  Methods in Artificial Intelligence, 2001.
[33]	K.C. Fukushima, “A Self-Organizing Multilayered Neural Network”, Bio-logical Cybernetics, 1975, pp.121–136.
[34]	L.E. Baum and T. Petrie, “Statistical Inference for Probabilistic Functions of Finite State Markov Chains”, Ann. Math. Stat., 1966, pp.1554–1563.
[35]	K. Jonsson, J. Matas, J. Kittler and Y. Li, “Learning support vectors for face verification and recognition”, IEEE, International Conference on Automatic Face and Gesture Recognition, 2000, pp.208 - 213. 
[36]	Bernd Heisele, Purdy Ho, and Tomaso Poggio, “Face recognition with 70 support vector machines: Global versus component-based approach”, IEEE, International Conference on Computer Vision, July 2001, pp.688 - 694.
[37]	Vladimir N. Vapnik, “The nature of statistical learning theory” , New York: Springer-Verlag, 1995.
[38]	J. T. Tou, R. C. Gonzalez, “Pattern recognition principles”, Addison-Wesley Publishing Company, 1974.
[39]	J.C. Dunn, “A Fuzzy Relative of The ISODATA Process and Its Use in Detecting Compact Well-Separated Cluster”, Journal of Cybernetics, vol. 3, no. 3, 1973, pp.32–57. 
[40]	Chih-Chung Chang and Chih-Jen Lin,”LIBSVM -- A Library for Support Vector Machines”, Network : http://www.csie.ntu.edu.tw/