臺灣博碩士論文加值系統

現今，身份辨識的方法有很多種，傳統的辨別方式如密碼會有遺忘、混淆甚至遭人臆測或偷取，為了彌補這類問題的產生，生物辨識的技術隨之被提出，以生物特徵辨識個人身份的種類，大致上可分為二種，第一種為生理特徵，例如：人臉、指紋、虹膜與掌紋等，第二種則是以行為特徵作為辨識身份，例如：語音、簽名或是走路的速率等。一般來說，以個人簽名作為身份辨識的方法是大眾廣為接受的。
本研究利用數位手寫板結合本論文所發展之特徵擷取系統，讓使用者能在書寫個人簽名時，記錄下使用者的簽名特徵，包含靜態特徵與動態特徵，靜態特徵如：起筆點X座標、起筆點Y座標、終筆點X座標、終筆點Y座標、總筆劃像素、框字長度、框字寬度、框字面積、總筆劃數，動態特徵如：總簽名時間、最後一筆劃速度以及簽名的書寫速率，透過收集真實與仿冒簽名的特徵資訊建立簽名資料庫，並應用倒傳遞類神經網路之學習架構，藉以判斷是否為本人簽名或是仿冒簽名。隨機仿簽測試實驗結果平均為真實簽名被拒絕率2.71%；仿冒簽名被接受率則為3.22%，此外，本研究進一步挑選一間餐廳的用餐者進行辨識率測試，其實驗結果真實簽名被拒絕率平均為0.00%；仿冒簽名被接受率平均為1.8%。最後，在技術仿簽實驗部分，本論文發現加入技術仿簽的訓練資料後，可百分之百有效拒絕冒名簽名。

In recent years, there are many ways to verify personal identity. Traditional methods is using password. However, it might be forgotten, lost, or stolen. To prevent occurring these uncomfortable situations, biometrical verifying identity methods is proposed. There are two categories. First method is physiological verification (e.g. face, fingerprints, iris, and hand). The other is behavioral verification (e.g. signature verification, speech, and walk speed. Recent researches show that signature verification system is welcome by most people.
In this study, I use WACOM digital tablets to collect information and bring it into signature verification system. It can record 12 signature characteristics of user. The static feature includes pen down position, pen up position, total pixels, etc. And, the dynamic feature includes overall signature time, speed, etc. We collect original signatures samples and forgery signatures samples to set signature database. The characteristics of database are used to Training samples. Training samples will calculate weights from back-propagation neural network. And the system will verify personal identify according to the weights. The experimental results show that the False Reject Rate is 2.71% and the False Accept Rate is 3.22% for random forgeries. Furthermore, we further choose several restaurant users to test data. The False Reject Rate is 0% and the False Accept Rate is 1.8% for the performance of the random forgeries detects experimental results. Finally, the False Accept Rate is 0% for skill forgeries after the proposed scheme utilizing the skill training data.

中文摘要 II
英文摘要 III
誌謝 IV
目錄 V
圖目錄 VII
表目錄 VIII
第一章 緒論 1
1.1研究動機 1
1.2研究目的 2
1.3研究架構 3
第二章 文獻探討 5
2.1手寫簽名的種類 5
2.2簽名特徵資訊的前處理 5
2.3簽名辨識方法 6
2.4小結 12
第三章 研究方法 13
3.1系統建構方法 13
3.2特徵選取 14
3.3簽名特徵資訊的前處理 20
3.4簽名辨識演算法 22
第四章 實驗與討論 28
4.1開發環境與系統需求 28
4.2簽名資料庫 29
4.3真實簽名樣本與仿簽樣本的學習和測試 30
4.4辨識率測試 34
4.4.1餐廳隨機仿簽與技術仿簽 37
第五章結論與未來展望 40
參考文獻 42

[1] A. I. Al-Shoshan, “Handwritten signature verification using image invariants and dynamic Features,” IEEE International Conference on Computer Graphics, Imaging and Visualisation, pp. 173-176, 2006.
[2] A. Kholmatov and B. Yanikoglu, “Identity authentication using improved online signature verification method,” Pattern Recognition Letters, Vol. 26 , pp. 2400-2408, 2005.
[3] A. Zimmer and L. Luan Ling, “A hybrid on/off line handwritten signature verification system,” Proceedings of the Seventh International Conference on Document Analysis and Recognition, pp. 424-428, 2003.
[4] A. K. Jain, F. D. Griess, and S. D. Connel, “On-line signature verification,” Pattern Recognition, Vol. 35, No. 12, pp. 2963-2972, 2002.
[5] B.Wirtz, “Stroke-based time warping for signature verification,” Proceedings International Conference on Document Analysis and Recognition, pp. 179-182, 1995.
[6] E. J.R. Justino, F. Bortolozzi, and R. Sabourin, “A comparison of SVM and HMM classifiers in the off-line signature verification,” Pattern Recognition Letters, Vol. 26, No. 9, pp. 1377-1385, 2005.
[7] F. R. Rioja, M. N. Miyatake, H. P. M., and K. T. M., “Dynamics features extraction for on-line signature verification,” Proceeding IEEE, Fourteen International Conference on Electronics, Communications and Computers, 2004.
[8] G. Elliott and N. Phillip, Mobile commerce and wireless computing system, AW, 2004.
[9] H. Lei, S. Palla, and V. Govindaraju, “ER2: An intuitive similarity measure for on-line signature verification,” Ninth International Workshop on Frontiers in Handwriting Recognition, pp. 191-195, 2004.
[10] H. Lei and V. Govindaraju, “A comparative study on the consistency of features in on-line signature verification,” Pattern Recognition letters, Vol. 26, No. 15, pp. 2483-2489, 2005.
[11] K. Huang and H. Yan, “On-line signature verification based on dynamic segmentation and global and local matching,” Optical Engineering, Vol. 34, No. 12, pp. 3480-3487, 1995.
[12] K. Huang and H. Yan, “Off-line signature verification based on geometric feature extraction and neural network classification,” Pattern Recognition, Vol. 30, No. 1, pp. 9-17, 1997.
[13] L. L. Lee, T. Berger, and E. Aviczer, “Reliable on-line human signature verification systems,” IEEE Transaction on Pattern Analysis and Machine Intelligence, Vol. 18, No. 6, pp. 643-647, 1996.
[14] M. Negnevitsky, “Artificial intelligence a guide to intelligent systems,” PEARSON Education, 2001.
[15] R. Martens, and L. Claesen, “On-line signature verification by Dynamic Time-Warping,” Proceedings of ICPR, 1996.
[16] T. Ohishi, Y. Komiya, and T. Matsumoto, “On-line signature verification using pen-position, pen-pressure and pen-inclination trajectories,” Proceeding IEEE International Conference on Pattern Recognition, pp. 4547-4550, 2000.
[17] 林政松，生物特徵辨識技術與印刷業之關係探討，中央印製廠，98期，2005。
[18] 張陽郎、方志鵬、許威烈、梁文耀、楊任傑、陳佳輝，利用GME演算法實作中文簽名辨識系統於PDA，2006北商學術論壇－資訊管理與實務研討會，2006。
[19] 許耀東、黃英軒、林宜樺，結合類神經網路與基因演算法應用於手寫簽名身份辨識系統，南台灣資訊科技與應用研討會，第18頁，2007。
[20] 陳建宏，結合實筆與虛筆資訊特徵作中文簽名確認，碩士論文，國立中央大學資訊工程研究所，桃園，2005。
[21] 詹焯燃，手寫數位簽名的生物特徵研究，碩士論文，私立東海大學資訊工程與科學研究所，台中，2006。
[22] 鄭弘裕，識別任意角度偏轉之手寫簽名，碩士論文，私立中華大學科技管理所，新竹，2004。
[23] 劉大仁，加速時間序列資料比對演算法設計及其在內容導向語音檢索系統之製作，碩士論文，國立高雄第一科技大學電腦與通訊工程系，高雄，2004。
[24] 葉怡成，類神經網路模式應用與實作，儒林圖書公司，2003。
[25] 賴承澤，利用虛筆特徵作中文簽名確認，碩士論文，國立中央大學資訊工程研究所，桃園，2001。