Title page for 80226001


[Back to Results | New Search]

Student Number 80226001
Author Chi-Jain Wen(溫志堅)
Author's Email Address No Public.
Statistics This thesis had been viewed 269 times. Download 10 times.
Department Optics and Photonics
Year 1995
Semester 2
Degree Ph.D.
Type of Document Doctoral Dissertation
Language zh-TW.Big5 Chinese
Title On-line Signature Verification System based on Wavelet Transform
Date of Defense
Page Count 164
Keyword
  • signature verification;wavelet transform;segmentation;feature extraction
  • Abstract In this thesis, we propose a new handwritten Chinese signature
    verification method based on segmentation and feature
    extraction using the multiresolution property of wavelet
    transform. One of the most important problem encountered in the
    process of signature verification is the segmentation of
    signature. As we know, there are informative characteristics
    hidden in the writing habit of an individual in signing his/her
    signatures, these features of writing habit exhibit themselves
    in the pen- down, pen-up and in the corner of the writing
    trajectory. Therefore, if we want to conserve the features
    hidden in the corners, the segmentation procedure for verifying
    a signature should be different from that for recognizing a
    character. In our verification method, the wavelet transform
    is employed to separate a signature into several segments for
    the purpose of preserving the corners of signature for local
    analysis. The multiresolution property of wavelet transform is
    also used to find the tremor features, which can be used to
    distinguish the forgery signatures from the genuine one. It
    has a high discriminating capability between the genuine
    signature and the more resembling forgeries. Some of features
    that are usually used in signature verification or character
    recognition, such as Fourier transform spectrum, velocity
    curve, signature profile and attributed string, are also
    employed to help to rejecting forgeries. Experimental results
    have shown that the new signature verificaiton method has high
    verifcation rate.
    Table of Content 封面
    摘要
    誌謝
    目錄
    圖目錄
    表目錄
    第一章 緒論
    1.1研究動機
    1.2專家鑑定的基本概念
    1.3國內外的研究概況
    1.4本論文的主要貢獻
    1.5論文內容概述
    本章參考文獻
    第二章 基本理論簡介
    2.1波元轉換理論
    2.1.1波元轉換基本理論
    2.1.2非正交波元轉換
    2.1.3離散波元轉換
    2.2動態時間歸正理論
    2.3相似度量測方法
    2.3.1歐氏距離
    2.3.2相關係數
    2.3.3屬性字串匹配
    本章參考文獻
    第三章 簽名資料處理
    3.1預處理
    3.1.1資料的讀取
    3.1.2雜點去除及座標平移
    3.1.3資料點插補及重新取點
    3.2簽名切割-由曲率轉折點切割
    3.2.1應用波元轉換搜尋曲率轉折點
    3.2.2尋找最佳匹配對應
    3.2.3最佳簽名切割探應
    3.3簽名切割-由角落切割
    3.3.1傳統的切割方式結合遞迴演算
    3.3.2利用波元轉換進行角落偵測
    3.3.3結合波元轉換與K-curvatrue的角落偵測
    本章參考文獻
    第四章 有效的特徵抽取
    4.1專家鑑定所使用的簽名特徵
    4.2整體特徵
    4.2.1整體特徵向量
    4.2.2座標位置
    4.2.3資料點方向角Chain-Code分布
    4.2.4傅氏轉換平面上的頻譜特徵
    4.3區域特徵
    4.3.1波元轉換平面上的特徵
    4.3.2速度曲線
    4.3.3筆段特徵向量字串
    4.3.4顫抖性特徵
    本章參考文獻
    第五章 簽名模板的建立
    5.1適當模板訓練資料組數的選取
    5.2單一模板的建立
    第六章 實驗結果分析比較
    6.1系統架構
    6.2單一特徵比對實驗結果
    6.3多層決策模式
    第七章 結論
    其他
    Reference
    Advisor
  • Hon-Fai Yau;Bor-Shenn Jeng(游漢輝;鄭伯順)
  • Files No Any Full Text File.
    Date of Submission

    [Back to Results | New Search]


    Browse | Search All Available ETDs

    If you have dissertation-related questions, please contact with the NCU library extension service section.
    Our service phone is (03)422-7151 Ext. 57407,E-mail is also welcomed.