The Scale Estimation Of Multi-Scale DEM Scaleless Representation

It is an important and frontier problem about multi-scale handling and representation of geo-spatial data in GIS. It has to resolve how to derive the different representations of the geo-spatial data because the different requirements of applications and analysis from users. At present, the multi-scale representations of DEM based on multi-resolution analysis (MRA) on wavelet transform has been studied much, and the methods of using them are successful. However, all of them did not give an analysis of the extent of generalization on theory. On the basis of scaleless transform of DEM data, this paper introduces a method of extracting stream network after the generalization, and the fractal transformation of square root model, which is the same with the quantity-fixed generalization model of the stream networks. Accordingly, every scale of DEM after transformation is deducted on theory.This paper is divided into five chapters and one brief introduction.The brief introduction describes the contexts about this paper and the general thoughts.The chapter one is introduction. It recites the application of wavelet analysis on multi-scale representation of DEM, indicates the study status in quo and limitation, and introduces the general aim and the main contents of this paper.The chapter two is the basic theory of wavelet analysis applying to multi-scale representation of DEM. It briefly reviewed the application and development history of wavelet analysis. This part focuses on wavelet multi-resolution analysis, Mallat algorithm, and actualizing method of multi-scale representation of DEM.The chapter three is the fractal transformation models of square root law. Firstly it introduces the basic theory of fractal, expatiates the limitations of square root model, and discusses fractal transformation models according to them. Based on these models, it describes the selection of them and methods of how to determine scaleless range.The chapter four is experiment. It introduces the extracting algorithm of stream networks and extracts them using ArcView software.The general applications of square root model, and the selection of quantitative model based on fractal analysis of stream networks are also introduced. Finally the producing data are used to the experiment and the analysis of the experiment results are given.The chapter five is conclusions and prospects. We summarize the whole paper and represent the issues to be studied in the future.