摘要
在同一地区,随着数字高程模型(DEM)分辨率的降低,或者地形图比例尺的缩小,DEM或地形图所描述的地形表面的细节部分不断舍弃而表现出宏观的骨架特征。本文将小波多尺度分析方法和方根模型结合模拟这一过程,由基于1:1万比例尺地形图建立的DEM生成了两种新的较小比例尺DEM,对比不同比例尺等高线可知,较小比例尺保持了较大比例尺的山体轮廓、山脊、谷地走向等地貌形态的塑造。采用坡度和剖面曲率两个参数及信息论的方法,分别对原始DEM和新生成的DEM进行分析和验证,结果表明,DEM经过数据综合,不但地形表面的轮廓越来越平缓,而且地形表面的细节也越来越平滑。利用方根模型作为小波高频系数阈值选取的依据更贴近于传统制图综合方法,能够将比例尺的改变与综合程度结合起来,实现任意比例尺DEM的自动综合。
In the same region, with reduction of DEM resolution or contraction of topographic map scale, the details of terrain described by DEM and topographic map are abandoned and only framework features are maintained. In this paper, the multi-scale wavelet analysis and Square Root Model is combined to simulate the general process. Firstly, the higher resolution DEM is decomposed into the low frequency and high frequency parts by Multi-scale wavelet. Secondly, according to the theory of Square Root Model, the threshold of wavelet coefficient in the high frequency is fixed by the ration of choice when the scale is transformed. Finally, all of the low frequency and high frequency is reconstructed. In older to analysis and verify the variety features of DEM, the study uses two important topographic factors slope and profile curvature, and views from Information Theory to comparative analysis. The result shows that the choice of threshold is close to the traditional cartographic generalization using the Square Root Model, and be able to join the scale variety with the degree of cartographic generalization, which realizes the automated generalization of random scale DEM.
出处
《测绘科学》
CSCD
北大核心
2008年第3期93-95,115,共4页
Science of Surveying and Mapping
基金
国家973重点基础研究发展计划项目(2007CB407203)
关键词
小波多尺度分析
数字高程模型
制图综合
坡度
剖面曲率
wavelet multi-scale analysis
digital elevation model（ DEM）
cartographic generalization
slope
profile curvature
