臺灣博碩士論文加值系統

本論文研究目的是人臉辨識系統，雖然人臉辨識系統已經發展多年了，許多人也都提出不同的方法，但是始終無法達到很好的效果，原因是有許多的因素無法克服。例如︰表情的變化、光的強弱、影像擷取設備特性的不同、人臉特徵的抽取等等。本論文的研究方向是朝彩色影像發展，分為三部分︰
第一部份為人臉偵測，由於HSI(Hue、Saturation、Intensity)色彩系統對明暗度的變化較不敏感，所以將RGB系統轉換為HSI系統，在彩色影像中的每一個像素都會對應到HSI平面的點。如果對應到HSI平面的點落在特定區域，我們就可以將那一個像素標誌成膚色像素。這個特定區域是統計座落在HSI平面上的膚色範圍而來的。再將膚色像素套用ISO DATA(Iterative Self-Organizing Data Analysis Technique Algorithm)把相近的顏色分群，個別的判斷每一群是否為人臉。判斷每一群是否為人臉就必須靠著器官的相對位置來判斷。
第二部分為特徵切割，要辨認不同的人，則必須要找出每個人唯一的特徵，然而要找出特徵就必須對影像做切割，如何判斷出哪一些區塊是我們所要的特徵(五官)，就必須靠著人臉固定不變的特徵(眼睛、鼻子、嘴巴)。本論文最後將眼睛、鼻子、嘴巴和整張臉的影像利用特徵空間投射(Eigenspace Projection)投射到特徵空間來當作每一個人的特徵值。
第三部份為認證系統，在本文中我們利用可塑性類神經網路(Plastic Perceptron Neural Network)，此網路在分類辨識上的應用特別適合，可平行分散處理不同的類別。當有新的類別要加入或替換時，不須全部網路重新學惜，與傳統的回傳式類神經網路比較，具有較高的可塑性。

This study proposes an face recognition system﹒This system has been developed many years﹒Many researcher also proposed many different method﹒Because there are many factors unable to overcome so recognition system have not a good result all the time﹒For example﹐the problem are facial expression﹐varying lighting﹐different quality of capture device﹐face feature extraction etc﹒The issue in this study is using color scenery images to develop recognition system﹒This system can be divided into three parts：
First part is face detection﹒Because HSI color system are not sensitive to the intensity variations﹒Hence﹐the RGB values of pixel in the input image are first transform into HSI color space﹒The every pixels in the image will be mapped onto one point in the HSI plane﹒If the corresponding point lies on the specified zone﹐then the pixel will be labeled as a skin pixel. The specified zone was statistic skin color range lies on the HSI space﹒ISO DATA(Iterative Self-Organizing Data Analysis Technique Algorithm)must be applied to separate the skin pixels into several clusters﹒We could exploit organ’s location on the face to decide every clusters whether was human face or not﹒
Second part is feature segment﹒In order to distinguish from different faces﹐we have to find out every unique face’s feature﹒We must segment image before find out feature﹒In order to find out which one are organs that we want﹒The invariable features(eye﹐nose﹐lip) on the face have to be exploited﹒In this thesis﹐“Eigenspace Projection”was applied to project eye﹐nose﹐lip and face’s image on the eigenspace﹐then many feature values are gotten﹒
Third part is verification system﹒This system is implemented based on the “Plastic Perceptron Neural Network”﹒This network is more suitable for classification especially and it can parallel and distributed process different class. Network has not overall retraining when you replace patterns or add new ones.“Plastic Perceptron Neural Network”has more elasticity than conventional“Black-Propagation Neural Network”.

頁次
中文摘要 i
英文摘要 ii
誌謝iv
目錄v
表目錄ix
圖目錄x
壹、緒論 1
　一、研究動機 1
　二、人臉辨識之系統架構 1
(一)影像擷取 2
(二) 影像切割 2
(三) 特徵抽取 2
(四) 辨識系統 2
　三、論文結構 3
貳、人臉偵測 4
　一、緒言 4
　二、膚色偵測 5
　三、色彩分群演算法(Color Clustering Algorithm)11
　(一) 問題描述13
　　(二) 分群(Clustering)15
　(三) 分裂(Split)17
　　(四) 合併(Merge)19
(五) 實驗結果21
四、人臉驗證22
(一) 侵蝕運算23
　　(二) 膨脹運算23
(三) 閉合運算25
參、特徵抽取29
　一、緒言29
(一) 區別性29
　　(二) 可靠性29
(三) 獨立性29
　二、五官定位30
(一) 眼睛定位31
　　(二) 嘴巴定位34
　　(三) 鼻子定位37
三、五官特徵抽取40
(一) 原始特徵空間投射法41
1. 產生特徵空間41
　　　2. 投射訓練影像42
　　　3. 投射待辨識影像到特徵空間42
　　
(二) 快速特徵空間投射法43
1. 中心化影像43
　　　2. 產生資料矩陣43
　　　3. 產生互變異數矩陣43
　　　4. 計算互變異數矩陣的特徵值和特徵向量43
　　　5. 計算資料矩陣的特徵向量43
　　　6. 排列特徵向量44
肆、辨識系統45
　一、緒言45
　　(一) 神經元的生物模型45
1. 細胞體(Soma)46
　　　2. 軸突(Axon)46
　　　3. 樹突(Dendrites)46
　　　4. 突觸(Synapse)46
　　(二) 神經元的運算模型47
(三) 認知器與倒傳式學習法48
1. 認知器48
　　　2. 倒傳式學習法50
二、可塑性認知網路(Plastic Perceptron Network)53
(一) 計算輸出值54
(二) 調整網路權重值56
伍、實驗結果與討論59
　一、五官幾何距離實驗結果61
　二、特徵空間投射法結果63
　三、結果討論70
陸、結論與未來研究方向71
　一、結論71
　二、未來研究方向71
參考文獻72

[1] 連國珍， 1999， 數位影像處理 ， 儒林出版
[2] 楊武智， 1994， 影像處理與辨認 ，　全華科技
[3] 繆紹綱﹐2000﹐數位影像處理-活用MATLAB，儒林出版。
[4] 周義昌、蔡玉娟、劉榮宜、李素玲、戴敏倫，類神經網路的新架構：可塑性認知網路，電信研究季刊, 第22卷第3期, 6月, 1992, pp. 291- 306
[5] Ming-Hsuan Yang﹐David Kriegman and Narendra Ahuja﹐”Detecting Faces in Images : A Survey”﹐IEEE Transactions on Pattern Analysis and Machine Intelligence﹐Vol. 24﹐No. 1﹐pp. 34-58﹐January 2002.
[6] Ing-Sheen Hsieh﹐Kuo-Chin and FanChiunhsium Lin﹐”A statistic approach to the detection of human faces in color nature scene”﹐Pattern Recongnition Society ﹐Vol. 35﹐pp. 1583-1596﹐Jun. 2002.
[7] Yao-Hong Tasi﹐Yea-Shuan Huang and Tomaso Poggio﹐”Fast Face Detection Using Components”﹐IPPR Conference on Computer Vision﹐Graphics and Image Processing﹐15th﹐2002.
[8] K. I. Kim﹐S. H. Park and H. J. Kim﹐”Kernel Principal Component Analysis for Texture Classification”﹐IEEE Signal Processing Letters﹐Vol. 8﹐No. 2﹐February 2001.
[9] Lawrence W. Cychosz and Jun Zhang﹐”A Face Recognition System Using Holistic Regional Feature Extraction”﹐IEEE Computer Society Conference on Computer Vision and Pattern Recognition﹐pp. 3431-3434﹐1995.
[10] Zhujie and Y. L. Yu﹐”Face Recognition with Eigenfaces”﹐The Hong Kong University of Science and Technology.
[11] N. Eveno﹐A. Caplier and P. Y. Coulon﹐”A New Color Transformation for Lips Segment” ﹐IEEE Multimedia Signal Processing﹐pp. 3-8﹐2001.
[12] Haiyuan Wu﹐Qian Chen and Masahiko Yachida﹐”Face Detection from Images Using a Fuzzy Pattern Matching Method”﹐IEEE Transactions on Pattern Analysis and Machine Intelligence﹐Vol. 21﹐No. 6﹐June 1999.
[13] Chengjun Liu and Harry Wechsler﹐”A Shape and Texture-Based Enhanced Fisher Classifier for Face Recognition”﹐IEEE Transaction on Image Processing﹐Vol. 10﹐No. 4﹐April 2001.
[14] Sheng-Wen Jeng﹐”Speed Up Shape Extraction of Eyes and Mouth in a Face by Pre-Location the Facial Feature Points”﹐IPPR Conference on Computer Vision﹐Graphics and Image Processing﹐15th﹐pp. 74-80﹐2002.
[15] Olugbenga Ayinde and Yee-Hong Yang﹐”Region-Based Face Detection”﹐Pattern Recognition﹐Vol. 35﹐pp. 2095-2107﹐September 2001.
[16] Lian Hock Koh﹐Surendra Ranganath amd Y. V. Venkatesh﹐”An Integrated Automatic Face Detection and Recognition System” ﹐Pattern Recognition﹐Vol. 35﹐pp. 1259-1273﹐June 2001.
[17] Marian Stewart Bartlett﹐Javier R. Movellan and Terrence J.Sejnowski﹐”Face Recognition by Independent Component Analysis”﹐IEEE Transactions on Neural Networks﹐Vol. 13﹐No. 6﹐November 2002.
[18] Rong-Qin Luo and Wei-Zhi Luo﹐”Human Face Detection and Recognition by Improved Gray Matching Based on Geometric Invariant Features”﹐ IPPR Conference on Computer Vision﹐Graphics and Image Processing﹐15th﹐pp. 56-65﹐2002.
[19] Wendy S. Yambor﹐Bruce A. Draper and J. Ross Beveridge﹐”Analyzing PCA-Based Face Recognition Algorithm : Eigenvector Selection and Distance Measures”﹐Computer Science Department Colorado State University Fort Collins﹐July﹐2000.
[20] Yuan Y. Tang﹐Yu Tao and Ernest C. M. Lam﹐”New Method for Feature Extraction Based on Fractal Behavior”﹐Pattern Recognition﹐Vol. 35﹐pp. 1071-1081﹐April 2002.
[21] Jian Yang and Jing-Yu Yang﹐”Why can LDA be Performed in PCA Transformed Space﹖”﹐ Pattern Recognition﹐Vol. 36﹐pp. 563-566﹐January 2001.