Scale Issues In Information Extraction From IKONOS Data

With the development and widespread use of remote sensing technologies, the remotely sensed imagery have been the main resource data for geographical phenomenon analysis. The spatial resolution is the most important and fundamental characteristics of remotely sensed images, namely measurement scale. Because of most geographical phenomenon are scale independent, the conclusions appropriate to one scale of environmental heterogeneity may be inappropriate when transferred to another scale.Scale variation and sensitivity have played an increasingly important role in the employment of remotely sensed imagery. Studying on the scale issue are not only useful for selecting a appropriate scale and reducing uncertainty during information extraction, but also useful for taking advantage of multi-scale information.This dissertation study focuses on scale issue including scaling, scale effect and optimal scale in information extraction from IKONOS imagery. A multi-scale method is developed to classify IKONOS imagery. The main contents that this research has made are summarized as follow:(1) Scale effect of DEM on orthorectification accuracyIKONOS raw images usually contain such significant distortions caused by terrain relief that they cannot be used directly with map base products in a geographic information system (GIS). Orthorectification was performed to remove the distortion and scale effect of DEM on orthorectification were evaluated. 38 ground control points (GCPs) and 25 independent check...