Font Size: a A A

Application Research Of Multi-scale Representation Based On Wavelet Transform In Guizhou Karst Region Fan Xvliang

Posted on:2016-02-23Degree:MasterType:Thesis
Country:ChinaCandidate:X L FanFull Text:PDF
GTID:2180330479955464Subject:Geodesy and Survey Engineering
Abstract/Summary:PDF Full Text Request
As the frame data of national spatial data infrastructure, the digital elevation model plays an important role in national economy and national defense construction. DEM express the continuous variation terrain surface by regular or irregular elevation sampling on surface, which will certainly be constrained by different kinds of scale. Many geographical phenomena and law will appear only in certain scale. DEM and terrain analysis have obvious scale dependency.As an important and basic terrain parameter, slope is being used as a key input for regional soil erosion modeling.Slope is usually extracted from Digital Elevation Model with a medium and lower resolution. With the decrease of DEM resolution, the slope will tend to be gentle, namely the attenuation phenomena. This kind of attenuation will lead to the fact that the slope extracted by low-or moderate-resolution can’t reflect rise and fall situation of terrain, affecting the accuracy of soil erosion evaluation. The slope from lower resolution DEM must be re-scaled to approximate higher resolution slope.Gui Zhou, the center of Southern Karst, the proportion of the rocky desertification area is the first in China, the phenomenon of soil erosion is serious. The results of terrain factor extraction based on different scale DEM data are also different. One of the important basic works of the soil erosion assessment is the analysis of the terrain factor of various scales.The Wavelet Multi-resolution and the multi-scale representation of DEM have surprising similarity. Based on wavelet transform theory, take for Huajiang basin sample area as an example, the multi-scale representation of DEM is realised. The main contents and results are as follows:(1)The determination of DEM optimum resolution. DEM optimum resolution is the critical value of size of mesh lattice units when the approximation level achieve the best for surface expression simulated by grid dem. DEM optimum resolution in experimental area can be determined by slope extracted by different kinds of resolution、roughness、ground curvature etc terrain factors. We draw a conclusion that 5m and 20 m can be used as the optimum resolution of 1:10000 and 1:50000 DEM production in experimental area.(2)The multi-scale representation of DEM in experimental area can be realised base on the multi-resolution technique by wavelet transform. Through expounding wavelet transform theory and the idea of wavelet multiresolution, the proper processing way of DEM boundaries are determined according to topographic structure and DEM data structure. The biorthogonal wavelet Bior5.5 is selected as wavelet basis function, the wavelet high-frequency coefficients is processed based on soft threshold function. On the basis of 1:10000 scale DEM, the DEM which has certain control basic and different resolution can be derived, which can better reflect the whole fluctuation status of terrain.(3)The slope change based on histogram matching. The slope can be extracted based on different resolutions, the slope value extracted by DEM 5m resolution is as the reference value. The values of slope are converted by other resolutions through construction models. The results show that the attenuating situation can be solved in some extent by slope scaling.
Keywords/Search Tags:DEM, Wavelet Transform, multi-scale representations, slope scaling
PDF Full Text Request
Related items