張璞曾臺灣大學：電機工程學研究所徐雍智Shi, Yung-ChihYung-ChihShi2007-11-262018-07-062007-11-262018-07-062005http://ntur.lib.ntu.edu.tw//handle/246246/53531健保IC卡已全面實施，對於人民的醫療需求帶來莫大的幫助。然而這一張關係到每個人隱私與重要醫療資訊的卡片儼然成為盜用的目標，因此，本研究提出一防盜的機制，結合人臉辨識與IC健保卡系統，當民眾到醫院去掛號，利用讀卡機讀取IC健保卡時，讀卡機上裝設的CCD即時將民眾的人臉影像拍下，經由偵測的方式將人臉特徵抓取出來，同時系統讀取卡片上的資料，根據卡片上存放人臉的特徵資料與即時拍攝的影像的特徵資料作比對，比對出確定無誤即能證明此張IC健保卡為本人所有而非別人所盜用。由於每個人都有一張自己獨特的臉孔，是別人無法效仿的，利用此一特性有效杜絕被盜用及仿冒的風險。IC cards for health insurance has been taken effect throughout Taiwan. This is medically beneficial to the people. However, this card is closely related to everyone’s privacy and to their personal important medical data, has been the target of piracy. Therefore, the purpose of this research is to offer a mechanism to prevent the occurrence of piracy by combining human face recognition method into the system for IC health insurance cards. When people register at hospital, their IC card is readed by card reader. At the same time ,the CCD equipped on the card reader will take pictures of their faces. From the features of the human face stored on the IC card, we can compare these with the instantaneous picture to make sure that the user of this IC card is the correct person .As everyone has his unique face features ,it is very difficult to imitate.We can use this characteristic to eliminate the risk of being counterfeited.第一章 序論…………………………………………………………………………1 1-1概論……………………………………………………………………………1 1-2 研究動機………………………………………………………………………3 1-3 智慧晶片IC卡………………………………………………………………3 1-4 系統目標………………………………………………………………………4 第二章 原理與方法…………………………………………………………………6 2-1 人臉偵測………………………………………………………………………6 2-2 人臉辨識………………………………………………………………………7 2.2.1 Geometric Feature-Based Methods…………………………………7 2.2.2 Template-Based Methods………………………………………………7 2.2.2.1 Karhunen-Loeve Expansion-Based Methods…………………8 2.2.2.2 Linear Discriminant-Based Methods…………………………8 2-3 IC卡的規格與架構……………………………………………………………9 2.3.1 Integrated Circuit Card (ICC)……………………………………10 2.3.2 Interface Device (IFD)……………………………………………11 2.3.3 Interface Device Handler (IFD Handler)………………………11 2.3.4 ICC Resource Manager………………………………………………12 2.3.5 Service Provider……………………………………………………12 2.3.5.1 ICC Service Provider…………………………………………12 2.3.5.2 Cryptographic Service Provider……………………………13 2.3.6 ICC-Aware Application………………………………………………13 第三章 系統架構…………………………………………………………………15 3-1 系統架構簡介………………………………………………………………15 3-2 人臉偵測……………………………………………………………………16 3.2.1 YUV色彩模型…………………………………………………………17 3.2.2 皮膚及嘴唇的色彩空間………………………………………………18 3.2.2.1 皮膚的色彩空間………………………………………………18 3.2.2.2 嘴唇的色彩空間………………………………………………18 3.2.3 人臉偵測架構…………………………………………………………20 3.2.3.1 偵測皮膚區域…………………………………………………20 3.2.3.2 偵測嘴唇區域…………………………………………………24 3-3 人眼與眉毛偵測……………………………………………………………27 3.3.1 Gradient Operator…………………………………………………27 3.3.2 Laplacian Operator…………………………………………………28 3.3.3 Averaging filter……………………………………………………29 3.3.4 人眼與眉毛偵測………………………………………………………30 3-4 人臉辨識……………………………………………………………………35 3.4.1 特徵臉演算法…………………………………………………………35 3.4.2 人臉辨識系統架構……………………………………………………35 3.4.3 特徵眼演算法…………………………………………………………40 3.4.4 亮度的補償……………………………………………………………41 第四章　實驗結果…………………………………………………………………44 4-1 攝影機………………………………………………………………………44 4-2 人臉與人眼眉毛偵測………………………………………………………47 4.2.1人臉偵測………………………………………………………………47 4.2.2 人眼與眉毛偵測………………………………………………………48 4-3 人臉辨識……………………………………………………………………51 4.3.1 實驗一…………………………………………………………………51 4.3.2 實驗二…………………………………………………………………52 4.3.3 實驗三…………………………………………………………………53 4.3.4 實驗四…………………………………………………………………53 4.3.5 實驗五…………………………………………………………………55 4-4 IC卡讀寫……………………………………………………………………56 第五章 討論與結論………………………………………………………………59 5-1 討論…………………………………………………………………………59 5-2 結論…………………………………………………………………………61 第六章　未來工作…………………………………………………………………62 參考文獻……………………………………………………………………………631923586 bytesapplication/pdfen-US嘴唇皮膚偵測邊緣偵測PCA演算法IC cardmouth and skin detectedge detectPCA algorism[SDGs]SDG3thesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/53531/1/ntu-94-R91921105-1.pdf