基于图像处理的笔迹防伪检测方法
发布时间：2021-11-23 00:17
在各种犯罪活动中文档伪造是一种常见的伪装手段,由于其易于实施而难于分辨,文档伪造的检测技术显得更加重要而迫切,通过对文档中笔迹改动的检测辨别文档真伪是其中重要的手段。然而,篡改者往往使用相近颜色的墨水对文档进行篡改,因此往往无法靠眼睛进行分辨。数字图像处理的方法提供更加精细的分辨能力,因此为这一问题带来新的解决方案。本文针对文档篡改中的笔迹检测问题,提出了两种基于图像处理的检测方法。论文的主要创新性工作包括:首先,综述并分析了用于笔记检测的重要图像特征,对比了不同的色彩空间表达。鉴于基于数字图像的笔迹检测算法目前缺少测试样本,我们首先建立了一个包含各种篡改情况的样本库,并通过扫描仪对样本进行数字化,得到了一个数字图像样本库,为后续的算法研究提供支撑;第二,提出了一种基于RGB特征距离的篡改笔迹检测方法,通过提取RGB通道的9个特征并计算特征与文档中所有目标的距离进行检测。首先文档被分割成独立的目标对象,然后计算每个目标对象的9个特征,以及用均方根误差表示的两两特征向量之间的距离。然后通过改进的Thompson Tau算法检测异常点并进行第二轮异常点检测。实验证明了所提出方法在检测和定位...
【文章来源】：哈尔滨工业大学黑龙江省 211工程院校 985工程院校
【文章页数】：89 页
【学位级别】：硕士
【文章目录】：
摘要
ABSTRACT
Chapter 1 General Introduction
1.1 What is the handwriting forgery
1.2 Background and history of handwriting forgery
1.2.1 Famous forgery cases of the 20th century
1.3 Features of handwriting
1.3.1 Characteristics of handwriting
1.4 Types of handwriting forgery
1.5 Research abroad and limitations
1.5.1 Analysis of literatures
1.6 Problem statement and motivation
1.7 Thesis outline and contributions
Chapter 2 Related Works
2.1 Color features
2.1.1 Color models
2.1.2 Color histogram
2.2 English handwriting dataset for forgery detection
2.2.1 Dataset description
2.3 Summary
Chapter 3 Handwriting Forgery Detection Based On Ink Color Features
3.1 Introduction
3.2 Proposed method
3.2.1 Pre-processing and segmentation
3.2.2 Features extraction
3.2.3 Distance measurement
3.2.4 Outliers detection
3.3 Experiment results
3.3.1 Detection performance
3.3.2 Experiment results and discussion
3.4 Summary
Chapter 4 Handwriting Forgery Detection Based On Structural Similarity Index of Color Histogram
4.1 Introduction
4.2 Proposed method
4.2.1 Pre-processing and segmentation
4.2.2 Color histogram
4.2.3 The structural similarity index measure
4.2.4 Similarity matching and forged object detection
4.3 Experiment results
4.3.1 Visual results
4.3.2 Detection performance
4.3.3 Running time
4.4 Summary
Conclusions
References
List of Publications
Acknowledgements
本文编号：3512730
【文章来源】：哈尔滨工业大学黑龙江省 211工程院校 985工程院校
【文章页数】：89 页
【学位级别】：硕士
【文章目录】：
摘要
ABSTRACT
Chapter 1 General Introduction
1.1 What is the handwriting forgery
1.2 Background and history of handwriting forgery
1.2.1 Famous forgery cases of the 20th century
1.3 Features of handwriting
1.3.1 Characteristics of handwriting
1.4 Types of handwriting forgery
1.5 Research abroad and limitations
1.5.1 Analysis of literatures
1.6 Problem statement and motivation
1.7 Thesis outline and contributions
Chapter 2 Related Works
2.1 Color features
2.1.1 Color models
2.1.2 Color histogram
2.2 English handwriting dataset for forgery detection
2.2.1 Dataset description
2.3 Summary
Chapter 3 Handwriting Forgery Detection Based On Ink Color Features
3.1 Introduction
3.2 Proposed method
3.2.1 Pre-processing and segmentation
3.2.2 Features extraction
3.2.3 Distance measurement
3.2.4 Outliers detection
3.3 Experiment results
3.3.1 Detection performance
3.3.2 Experiment results and discussion
3.4 Summary
Chapter 4 Handwriting Forgery Detection Based On Structural Similarity Index of Color Histogram
4.1 Introduction
4.2 Proposed method
4.2.1 Pre-processing and segmentation
4.2.2 Color histogram
4.2.3 The structural similarity index measure
4.2.4 Similarity matching and forged object detection
4.3 Experiment results
4.3.1 Visual results
4.3.2 Detection performance
4.3.3 Running time
4.4 Summary
Conclusions
References
List of Publications
Acknowledgements
本文编号：3512730
本文链接：https://www.wllwen.com/shekelunwen/gongan/3512730.html
