Analysis of Multiresolution Data Fusion Techniques
Duane B. Carter
Thesis submitted to the Faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of Master of Science in Geography
Dr. James B. Campbell, Chair
Dr. John Randolph
Dr. Randolph H. Wynne
March 23, 1998
Blacksburg, Virginia
ABSTRACT
In recent years, as the availability of remote sensing imagery of varying resolution has increased, merging images of differing spatial resolution has become a significant operation in the field of digital remote sensing. This practice, known as data fusion, is designed to enhance the spatial resolution of multispectral images by merging a relatively coarse-resolution image with a higher resolution panchromatic image of the same geographic area. This study examines properties of fused images and their ability to preserve the spectral integrity of the original image. It analyzes five current data fusion techniques as applied to three complex scenes to assess their performance. The five data fusion models used include one spatial domain model (High-Pass Filter), two algebraic models (Multiplicative and Brovey Transform), and two spectral domain models (Principal Components Transform and Intensity-Hue-Saturation). SPOT data were chosen for both the panchromatic and multispectral data sets. These data sets were chosen for the high spatial resolution of the panchromatic (10 meters) data, the relatively high spectral resolution of the multispectral data, and the low spatial resolution ratio of two to one (2:1). After the application of the data fusion techniques, each merged image was analyzed statistically, graphically, and for increased photointerpretive potential as compared with the original multispectral images. While all of the data fusion models distorted the original multispectral imagery to an extent, both the Intensity-HueSaturation Model and the High-Pass Filter model maintained the original qualities of the multispectral imagery at the highest level. The High-Pass Filter model, designed to highlight the high frequency spatial information , provided the most noticeable increase in spatial resolution.
Table of Contents
Abstract ii
Acknowledgments iii
List of Figures v
List of Tables vii
Chapter One: Introduction 1
Spatial and Spectral resolution 1
The Factor of Scale 2
Advantages of Data Fusion 2
Applications 3
Data Characteristics 5
Test Area’s 6
Research Goals 7
Chapter Two: Data Fusion Models 8
Rectification and Registration 8
High-Pass Filter Model 9
Multiplicative Model 10
Brovey Transform 10
Principal Components Transformation 10
Intensity-Hue-Saturation 12
Chapter Three: Methods of Evaluation 15
Methods of Evaluation 15
Chapter Four: Results and Analysis 18
Discussion
Chapter Five: Conclusion 38
Appendix A: F Test Results 42
Appendix B: Photointerpretive Evaluation 45
Appendix C: Results of Photointerpretive Evaluation 46
Appendix D: Results of ANOVA: two way without replication 49
Bibliography 51
Vita 52
Conclusion
While each of the data fusion techniques analyzed here enhanced the spatial resolution of the original XS image, none maintained complete spectral integrity. This is not to say that the application of data fusion to disciplines requiring high spatial integrity cannot benefit from the use of the models, specifically the IHS and the HPF model, it is to say that the analyst should be aware of the benefits as well as the spectral distortion caused by applying the techniques. When spectral integrity cannot be lost then the analyst should regard the gain in spatial resolution provided by data fusion as nil. Although, the spatial enhancements provided with a correlated spectral scene can often be of more benefit than that of a relatively low spatial resolution XS image. Thus, the use of data fusion techniques, when two complementary multispatial scenes are available, should be considered as a potential tool for all analysts.
Full Text: pdf
ليست هناك تعليقات:
إرسال تعليق