臺灣博碩士論文加值系統

本論文目的在建立一個臉部驗證系統，可由輸入一段臉部動態影確認身份。此驗證系統主要分成三個部份，第一部份去除不需要的背景資訊與抓取臉部區域，利用膚色與視訊片段的畫面差異來達到目標，其中膚色模組的建立不須經由額外資料的訓練，可由輸入影像自動產生；第二部份為臉部特徵點擷取，包含了如何找出臉部各個特徵的大概位置，並且引用AAM的方法找出較準確的對應點，其可利用額外資料訓練的資訊來處理亮度變化及三維空間變形的影響；第三部份為臉部確認，在未校正的情況下建立測試與資料庫的三度空間位置與兩者之間轉換，使用測試三維特徵點投影到資料庫二維影像上作確認。

The system proposed in this thesis uses a face motion video clip to perform face verification. This system consists of three parts.The first part is to separate the background from the face by using the frame difference technique and the imformation of skin color .This skin color model is constructed automatically from the training input viedo clip. The second step is to extract the facial feature points based on the AAM shape and appearance models built from the training image set .This method can resist the intensity change and image geometric variation.The final part is to verify the identity of the face.It reconstructs the 3D model without camera calibration and verifies the face identity by registering the facial feature points to the gallery face image through the 2D image projection of the 3D face model constructed.

中文摘要................................iii
英文摘要................................iv
目錄................................vi
表目錄................................vii
圖目錄................................viii
第1章　緒論 1
1.1 研究動機與目標 1
1.2 相關研究 1
1.3 研究方法介紹 2
1.4 論文組織 3
第二章　臉部區域的Segmentation 4
2.1 利用Frame difference，去除大部份不必要的背景 4
2.1.1 相鄰的video clips的選取 4
2.1.2 Frame difference計算 4
2.1.3 Frame difference較強且集中區域的偵測 5
2.2 利用Hue與Saturation來抓取皮膚所包含的區域。 5
2.2.1 建立Hue與Saturation之二維color space 5
2.2.2分類皮膚區域的樣本選取 6
2.2.3 Maximum Likelihood建立樣本區域的Mixture Gaussian Model 6
2.2.4 相似度計算與皮膚區域擷取 8
第三章　臉部特徵區域偵測 10
3.1 眼珠偵測與嘴唇偵測 10
3.1.1 眼珠偵測 10
3.1.2 嘴唇偵測 11
3.2 眉毛與耳朵區域抓取 13
3.2.1 眼珠連線及距離計算與產生眼睛、嘴巴區域 13
3.2.2 眉毛區域抓取 13
3.2.3 耳朵區域抓取 14
第四章　AAM臉部特徵邊緣線偵側與特徵點擷取 16
4.1 AAM概述 16
4.1.1 AAM Training Step 16
4.1.2 AAM Searching Step 18
4.2 AAM Data Base的建立 18
4.2.1 AAM Data Base的影像拍攝 19
4.2.2 定義臉部特徵的重要特徵點 20
4.2.3 建立臉部特徵的AAM Data Base 20
4.3 臉部特徵區域的初步fitting – PSO 23
4.3.1 PSO簡介 23
4.3.2 PSO演算法與參數最佳化 23
4.3.3 使用model之mean shape 初步對應以降低rotation、scaling影響 24
4.4 臉部特徵邊緣線偵測與特徵點擷取 24
第五章 Bilinear+LMeds建立3D Model與臉部辨識 26
5.1 建立3D Model 26
5.1.1 Bilinear Method 概述 26
5.1.2 LMeds Method 概述 27
5.1.3 Data Condition概述 28
5.2 Uncalibration情況的臉部辨識 30
5.2.1 透視投影空間轉換 30
5.2.2 投影至2D影像上的臉部辨識與分析 31
第六章 實驗與臉部辨識結果 32
6.1 拍攝資料建立 32
6.2 臉部皮膚區域抓取結果 35
6.3 臉部特徵區域抓取與重要特徵點擷取結果 39
6.3.1 AAM Data Base 建立 39
6.3.3 臉部特徵點擷取流程範例 41
6.3.4 測試資料特徵點之代表性結果 44
6.4 臉部辨識結果 46
6.4.1 B為測試對象比對結果 46
6.4.2 A為測試對象比對結果 48
6.4.3 C為測試對象比對結果 49
6.5 比對結果總整理 51
第七章 結論與未來工作 52
7.1 結論 52
7.2 未來發展方向 52
References 53

[1] Rein-Lien Hsu, Mohamed Abdel-Mottaleb ,Anil K. Jain, “Face Detection in Color Images”, PATTERN ANALYSIS AND MACHINE INTELLIGENCE Vol 24, pp. 696-706, 2002
[2] Hayit Greenspan, Jacob Goldberger, Itay Eshet, “Mixture model for face-color modeling and segmentation”. Pattern Recognition Letters, vol. 22, pp. 1525-1536, Sept. 2001
[3] Christophe Garcia, Georgios Tziritas ,“Face Detection Using Quantized Skin Color Regions Merging and Wavelet Packet Analysis” Trans. IEEE on MULTIMEDIA, VOL. 1,pp. 264 - 277 , Sept. 1999
[4] Hongliang Li, King N. Ngan, Qiang Liu, “FaceSeg: Automatic Face Segmentation for Real-Time Video”Trans. IEEE on MULTIMEDIA, VOL. 11, pp. 77-88 , Jan. 2009
[5] Liya Ding, Aleix M. Martinez, “Precise Detailed Detection of Faces and Facial Features” Dept. Electrical and Computer Engineering The Ohio State University, Columbus, pp. 77-88 June Aug. 2008
[6] Timothy F. Cootes, Gareth J. Edwards, and Christopher J. Taylor , “Active Appearance Models” Trans. IEEE on PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. 23, pp. 681-685 , JUNE 2001
[7] Lung-Kai Cheng,” Face Recognition Using 2D/3D Features” Master thesis, Department of Computer Science, National Chiao Tung University, August 2006.
[8] Graham D. Finlayson, Bernt Schiele, James L. Crowley “Comprehensive Colour Image Normalization” Proc. European Conference on Computer Vision Vol I, LNCS 1406, pp. 475-490,1998.
[9] Shyjan Mahamud, Martial Hebert, Yasuhiro Omori*, Jean Ponce “Provably-Convergent Iterative Methods for Projective Structure from Motion” Proc. IEEE Conference on Computer Vision and Pattern Recognition Vol 1, pp. 1018-1025 2001
[10] Chiao-Wen Cheng “Accurate Camera Pose Estimation from the Noisy Calibrated Stereo Images Based on LMeds Initial Results” Master thesis, Department of Computer Science, National Chiao Tung University, August 2008
[11] Marian Stewart Bartlett, Javier R. Movellan, and Terrence J. Sejnowski, “Face Recognition by Independent Component Analysis” Trans. IEEE on Neural Networks, VOL. 13,pp. 1450-1464, Nov. 2002