臺灣博碩士論文加值系統

本論文提出一個以特徵點排除(keypoint exclusion)為基礎並結合雙重特徵點偵測之人臉辨識方法。傳統之尺度不變特徵變換(Scale Invariant Feature Transform, 簡稱SIFT) 有下列三個缺點: (1)只單使用一種特徵偵測方法，在尺寸較小的影像中得到的特徵點個數可能過少，容易造成影像比對上的困難；(2)每一個測試影像之特徵點需要跟所有訓練範本的特徵點做比對，所花費的時間較長; (3)只有比較描述元(descriptor)之相似性，容易導致大量錯誤匹配(false match)的發生。
在特徵點之個數提升方面，我們提出結合SIFT與FAST (Features from accelerated segment test) 兩種不同類型的特徵點來做人臉影像之比對，由於FAST偵測法沒有對應之描述元，我們在影像中的每一個FAST特徵點上利用LOG (Laplace of Gaussian)函數做自動尺度選擇，找出較為適當之尺度，並計算出該點對應之SIFT描述元。另一方面，基於人臉影像中五官位置之相似性，我們提出三種特徵點排除的機制，透過相對位置、特徵點方向、尺度大小做初步的特徵排除，排除不可能之特徵點，以減少整體描述元之比對次數，因為使用排除的機制，不需要逐一點做比對，如此可節省大量的比對時間，執行不相關特徵點之初步排除，一來可減少錯誤匹配之情形發生，提高辨識之成功率，二來可縮短比對所需時間。
實驗部分使用ORL與YALE人臉資料庫，各取10人，每人10張影，相較於傳統之方法，本文所提出之三種特徵點排除法，均有較佳之辨識率，平均約提升5%，而運算時間減少一半以上，由實驗結果顯示，本論文所提出以角落與斑點特徵偵測的人臉辨識系統能達到較好的辨識結果以及較快之運算速度。

This thesis presents a face recognition system based on keypoint exclusion and dual keypoiont detection. There are three major problems with conventional SIFT (Scale Invariant Feature Transform). (1) It uses single type keypoint detector. For images of small size the number of detected keypoints may be too small and this causes difficulties on image matching. (2) Each keypoint of the test image is matched independently against all keypoints of the training images. This is very time consuming. (3) Only similarities between descriptors are compared and this may still causes some false matches.
To increase the number of keypoints, SIFT and FAST (Features from accelerated segment test) keypoints are combined for face image matching. Since there is no corresponding descriptor for FAST detector, the LOG (Laplace of Gaussian) function with Automatic Scale Selection is applied on each FAST keypoint to find proper scales and corresponding SIFT descriptors. On the other hand, based on the similarities between locations of features on human faces, three keypoint exclusion methods (relative location, orientation, and scale) are proposed to eliminate impossible keypoints for further descriptor matching. In this way, the number of false matches can be reduced and hence higher recognition rates can be obtained. On the other hand, matching time can also be reduced.
The proposed algorithms are evaluated with the ORL and the Yale face databases. Each database pick 10 person, every person get 10 image. Our proposed method shows significantly improvements on recognition rates over conventional methods.

摘要 I
ABSTRACT III
致謝 V
目錄 VI
圖目錄 VII
表目錄 IX
第一章 緒論 1
第一節 研究動機與目的 1
第二節 文獻回顧 2
第三節 論文架構 4
第二章 人臉偵測與辨識之相關方法 5
第一節 VIOLA AND JONES 人臉偵測器 5
一、積分影像 (Integral Image) 5
二、矩形特徵、弱分類器、Adaboost演算法 6
三、層疊分類器 7
第二節 尺度不變之特徵轉換(SIFT) 7
第三節 HARRIS-LAPLACE 角點偵測器 12
第四節 FAST 角落偵測法 13
第五節 人臉辨識相關方法 15
第三章 角落與斑點特徵結合與特徵排除 19
第一節 VIOLA & JONES人臉偵測器 20
第二節 FAST 角點偵測與SIFT描述元之提取 21
第三節 斑點與角落特徵點之結合 22
第四節 特徵點排除法(KEYPOINT EXCLUSION) 24
一、相對位置排除法(Relative Location exclusion Method) 25
二、相對方向排除法(Relative Orientation exclusion Method) 27
三、相對尺度排除法(Relative Scale exclusion Method) 29
第五節 匹配分數之計算 31
第四章 實驗結果與分析 33
第一節 人臉資料庫 33
第二節 實驗結果 34
第三節 時間比較 43
第五章 結論與討論 47
文獻參考 48

[1]Aly, Mohamed. "Face recognition using SIFT features." CNS/Bi/EE report 186 (2006).
[2]Lowe, David G. "Object recognition from local scale-invariant features." Computer vision, 1999. The proceedings of the seventh IEEE international conference on. Vol. 2 1999.
[3]Lowe, David G. "Distinctive image features from scale-invariant keypoints."International journal of computer vision 60.2 (2004): 91-110.
[4]Wang, Chieh-Chih, and Ko-Chih Wang. "Hand Posture recognition using Adaboost with SIFT for human robot interaction." Recent progress in robotics: viable robotic service to human. Springer Berlin Heidelberg, 2008. 317-329.
[5]Huang, Han-che. "The Study of Aerial Imageries Stitching Based on SIFT Algorithm." National Sun Vat-sen University Department of Marine Environment and Engineering Master''s thesis (2009).
[6]Zhou, Huiyu, Yuan Yuan, and Chunmei Shi. "Object tracking using SIFT features and mean shift." Computer vision and image understanding 113.3 (2009): 345-352.
[7]Bicego, Manuele, et al. "On the use of SIFT features for face authentication." IEEE Conference on Computer Vision and Pattern Recognition Workshop, CVPRW''06., 2006.
[8]Luo, Jun, et al. "Person-specific SIFT features for face recognition." IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP Vol. 2, 2007.
[9]D. R. Kisku, M. Tistarelli, J. K. Sing, and P. Gupta, "Face recognition by fusion of local and global matching scores using DS theory: An evaluation with uni-classifier and multi-classifier paradigm", IEEE Computer Vision and Pattern Recognition (CVPR) Workshop on Biometrics, 2009.
[10]Križaj, J., V. Štruc, and N. Pavešić. "Adaptation of SIFT features for face recognition under varying illumination." MIPRO, 2010 Proceedings of the 33rd International Convention. IEEE, 2010.
[11]Križaj, Janez, Vitomir Štruc, and Nikola Pavešić. "Adaptation of SIFT features for robust face recognition." Image Analysis and Recognition. Springer Berlin Heidelberg, 2010. 394-404.
[12]Han Yanbin, Yin Jianqin, Li Jinping, Human Face Feature Extraction and Recognition Base on SIFT, International Symposium on Computer Science and Computational Technology, vol. 1, pp.719-722, 2008.
[13]Paul Viola, Michael Jones, "Robust Real-time Object Detection", Second International Workshop on Statistical and Computational Theories of Vision - Modeling, Learning, Computing, and Sampling, 2001.
[14]Liu, T., Kim, S.-H., Lee, H.-S., Kim, H.-H.: Face Recognition base on a New Design of Classifier with SIFT keypoints. In: Intelligent Computing and Intelligent Systems, pp. 366–370, 2009.
[15] C. Fernandez and M. Vicente” Face recognition using multiple interest point detectors and SIFT descriptors", Proc. Int. Conf. Automatic Face Gesture Recognit , pp.1 -7 2008.
[16]Bay, Herbert, Tinne Tuytelaars, and Luc Van Gool. "Surf: Speeded up robust features." Computer Vision–ECCV 2006. Springer Berlin Heidelberg, 404-417, 2006.
[17]Juan, Luo, and Oubong Gwun. "A comparison of sift, pca-sift and surf." International Journal of Image Processing (IJIP) 3.4 (2009): 143-152.
[18]梁振升，「以OpenCV實現即時人臉偵測與辨識系統」，私立銘傳大學，電腦與通訊工程學系碩士論文，指導教授：李棟良，2012年。
[19]賴昆煒，「基於Radon轉換之人臉辨識系統」，私立銘傳大學，電腦與通訊工程學系碩士論文，指導教授：李棟良，2010年。
[20]Jen-Sheng Liang, Dong-Liang Lee. “A face detection and recognition system based on rectangular feature orientation” 2010 International Conference on System Science and Engineering (ICSSE), 2010. 495 – 499
[21]http://faculty.csie.ntust.edu.tw/~klchung/Prof_book_IPCV_ppt.htm
[22]Harris, Chris, and Mike Stephens. "A combined corner and edge detector." Alvey vision conference. Vol. 15. 1988.
[23]Mikolajczyk, Krystian, and Cordelia Schmid. "Scale & affine invariant interest point detectors." International journal of computer vision 60.1 (2004): 63-86.
[24]Mikolajczyk, Krystian, and Cordelia Schmid. "A performance evaluation of local descriptors." IEEE Transactions on Pattern Analysis and Machine Intelligence, 27.10 (2005): 1615-1630.
[25]Rosten, Edward, and Tom Drummond. "Machine learning for high-speed corner detection." Computer Vision–ECCV 2006. Springer Berlin Heidelberg, 430-443, 2006.
[26]Ghadarghadar, Nastaran, et al. "A SIFT-point distribution-based method for head pose estimation." IEEE International Workshop on Machine Learning for Signal Processing (MLSP), 2012.
[27]Ge, Wei, et al. "Recognition of Expression-Variant Faces Using SIFT Method." 2012 Fourth International Conference on Computational Intelligence and Communication Networks (CICN), 2012.
[28]Andersen, Vedrana, Lars Pellarin, and Renee Anderson. "Scale-Invariant Feature Transform (SIFT): Performance and Application.", 2006.
[29]Pirzada, S. J. H., et al. "A new adaptive threshold technique for improved matching in SIFT." Circuits and Systems (MWSCAS), 2011 IEEE 54th International Midwest Symposium on. IEEE, 2011.
[30]Alhwarin, Faraj, et al. "Improved SIFT-Features Matching for Object Recognition." BCS Int. Acad. Conf. 2008.
[31]Majumdar, Angshul, and Rabab Kreidieh Ward. "Discriminative SIFT features for face recognition." Canadian Conference on Electrical and Computer Engineering, 2009. CCECE''09.
[32]http://www.cl.cam.ac.uk/research/dtg/attarchive/facedatabase.html
[33]http://vision.ucsd.edu/content/yale-face-database
[34]http://www.robots.ox.ac.uk/~vedaldi/code/sift.html