| Chapter 1 Introduction | 第1-15页 |
| 1．1 Problem context and research motivations | 第10-12页 |
| 1．2 Research questions | 第12-13页 |
| 1．3 Objectives | 第13-14页 |
| 1．4 Synopsis of the thesis | 第14-15页 |
| Chapter 2 Conceptual framework | 第15-27页 |
| 2．1 Energy source for remote sensing | 第15-18页 |
| 2．1．1 Image space | 第16-17页 |
| 2．1．2 Feature space | 第17-18页 |
| 2．2 Digital image classification | 第18-19页 |
| 2．3 Unsupervised classification | 第19页 |
| 2．4 Supervised classification | 第19-25页 |
| 2．4．1 Training stage | 第20页 |
| 2．4．2 Classification stage | 第20-23页 |
| 2．4．2．1 Minimum distance to mean classifier | 第21页 |
| 2．4．2．2 Parallelepiped classifier | 第21-22页 |
| 2．4．2．3 Maximum likelihood classifier | 第22-23页 |
| 2．4．3 Image classification accuracy assessment | 第23-25页 |
| 2．4．3．1 The error matrix | 第24-25页 |
| 2．4．3．1．1 Overall accuracy | 第24-25页 |
| 2．4．3．1．2 Producer's accuracy | 第25页 |
| 2．4．3．1．3 User's accuracy | 第25页 |
| 2．4．3．2 Kappa statistics | 第25页 |
| 2．5 Summary | 第25-27页 |
| Chapter 3 Study area, data and data pre-preparation | 第27-36页 |
| 3．1 Study area | 第27-28页 |
| 3．2 Data | 第28-32页 |
| 3．2．1 Earth observation data | 第28-30页 |
| 3．2．2 Field survey data | 第30-32页 |
| 3．3 Data preparation | 第32-34页 |
| 3．3．1 Geo-reference of the imported ASTER data | 第32页 |
| 3．3．2 Geometric correction of remote sensing data | 第32-33页 |
| 3．3．3 Atmospheric correction | 第33-34页 |
| 3．3．4 Refinement of the field survey data | 第34页 |
| 3．4 Methodology to perform this research | 第34页 |
| 3．5 Summary | 第34-36页 |
| Chapter 4 Unsupervised classification results | 第36-45页 |
| 4．1 Unsupervised classification | 第36页 |
| 4．2 Algorithm used in unsupervised ASTER and LANDSAT ETM+ classification | 第36-37页 |
| 4．3 Initial clustering options setting for unsupervised classification | 第37页 |
| 4．4 Unsupervised ASTER classification | 第37-40页 |
| 4．4．1 Procedure and result of unsupervised classification with ASTER image | 第37-38页 |
| 4．4．2 Accuracy assessment of the classification result | 第38-40页 |
| 4．5 Unsupervised Landsat ETM+ classification | 第40-43页 |
| 4．5．1 The process of unsupervised classification for Landsat ETM+ classification | 第41页 |
| 4．5．2 Accuracy assessment of the unsupervised Landsat ETM+ classification | 第41-43页 |
| 4．6 Discussion of this chapter | 第43页 |
| 4．7 Summary | 第43-45页 |
| Chapter 5 Results of supervised classification | 第45-64页 |
| 5．1 Training samples selection and evaluation | 第45-46页 |
| 5．2 Classification algorithms selection | 第46页 |
| 5．3 Accuracy assessment | 第46-47页 |
| 5．4 ASTER data preparation before classification | 第47-49页 |
| 5．5 Supervised ASTER （VNIR+SWIR） classification result | 第49-54页 |
| 5．5．1 Accuracy assessment by error matrix formed by random sampling | 第51-52页 |
| 5．5．2 Accuracy assessment by Kappa statistics | 第52-53页 |
| 5．5．3 Accuracy assessment by error matrix formed by test areas | 第53-54页 |
| 5．6 Supervised ASTER （VNIR+TIR） classification result | 第54-58页 |
| 5．6．1 Accuacy assessment by error matrix formed by random sampling | 第55-56页 |
| 5．6．2 Accuracy assessment by Kappa statistics | 第56-57页 |
| 5．6．3 Accuracy assessment using error matrix formed by test areas | 第57-58页 |
| 5．7 Supervised classification result of Landsat ETM+ | 第58-62页 |
| 5．7．1 Accuacy assessment by error matrix formed by random sampling | 第60-61页 |
| 5．7．2 Accuracy assessment by Kappa statistics | 第61页 |
| 5．7．3 Accuracy assessment using error matrix formed by test areas | 第61-62页 |
| 5．8 Summary | 第62-64页 |
| Chapter 6 GIS manipulation to improve the classification results | 第64-70页 |
| 6．1 Model design to improve the classification results | 第65-67页 |
| 6．2 Final classified image combining the extracted classes | 第67-68页 |
| 6．3 Accuracy assessment by error matrix formed using test areas as reference data | 第68-69页 |
| 6．4 Summary | 第69-70页 |
| Chapter 7 Conclusions and Recommendations | 第70-73页 |
| 7．1 Conclusions | 第70页 |
| 7．2 Data and Methodology | 第70-71页 |
| 7．3 Limitation of this research | 第71-72页 |
| 7．4 Recommendation | 第72-73页 |
| Reference | 第73页 |
