Research on Image Classification And Fusion Based on Machine Learning Techniques

目录

声明

Abstract

摘要

Publications

Abbreviations

Contents

List of Figures

List of Tables

CHAPTER 1 INTRODUCTION

1．1 Introduction

1．1．2 Multi-Focus Image Fusion

1．1．3 Hyperspeetral Images Classification

1．2 Thesis Outline

CHAPTER 2 PAN-SHARPENING

2．2 Image Fusion Applications

2．3 Remote Sensing Applications

2．4 Panchromatic and Multi-spectral Imaging

2．4．1 Panchromatic Image

2．4．2 Multi-spectral Image

2．5 Classification of Pan-Sharpening Techniques(Literature review)

2．5．1 Component Substitution(CS)Techniques

2．5．2 Restoration Based(RB)Techniques

2．5．3 Multi-Resolution Analysis(MRA)Techniques

2．6 Research Objectives

2．7 Contributions

2．8 Flow of Work(How the whole work in process)

2．9 Proposed Scheme

2．10 Same domain

2．10．1 Laplacian Filter(LF)

2．10．2 Unsharp Mask

2．11 Cross domain(LF+DFT)

2．12 Matting Model(MM)Pan-sharpening

2．13 Frequency Components

2．14 Motivations of Using LF+DFT

2．15 Why Cross-domain?

2．16 Experiments

2．16．1 Dataset

2．16．2 Results and Discussion

2．16．3 Qualitative Analysis

2．16．4 Quantitative Analysis

2．17 Conclusion

2．18 Future Work

CHAPTER 3 FUSION ON MULTI-FOCUS IMAGES

3．1．2 Domains of Image Fusion

3．1．3 Levels of Image Fusion

3．2 Proposed Scheme

3．2．1 Stationary Wavelet Transform(SWT)

3．3 Experiments

3．3．1 Qualitative Performance Metrics

3．3．2 Datasets

3．3．3 Experimental Results and Discussions

3．4 Conclusion

CHAPTER 4 HYPERSPECTRAL IMAGE(HSI)CLASSIFICATION

4．1 Introduction

4．2 Typical NRS Classifier

4．3 Proposed Scheme

4．4 Manhattan Distance

4．5 Motivation for Using Manhattan Distance

4．6 Analysis of the Proposed NRS-MD

4．7 Experiments

4．7．1 HSI Datasets

4．8 Parameter Tuning

4．9 Experimental Results and Analyses

4．10 Conclusion

CHAPTER 5 CONCLUSION

5．1 Conclusion

References