基于特征的地形简化中的不确定性研究

发布时间：2018-02-03 06:26

本文关键词： 数字高程模型地形简化不确定性　出处：《西南交通大学》2017年硕士论文　论文类型：学位论文

【摘要】：数字高程模型是重要的空间数据,也是地形分析的主要数据来源。尺度效应导致同一地形在不同尺度下具有不同的地形特征,地形的表达也不尽相同。一般的,在地形的多尺度建模中,较粗糙的数字高程模型来源于对精细尺度的数字高程模型的信息筛减和地形重建,这一过程称为地形简化。影响数字高程模型精度的因子有很多种,如地形地貌、采样间隔、插值方法等等。无论是建立数字高程模型还是基于数字高程模型进行地形分析,熟知这些因子的影响是十分重要的。在对地形进行多尺度建模的过程中,依旧需要考虑这些因子的影响。已有学者对地形地貌、采样间隔、插值方法和空间分辨率等因子进行了研究,本文则着眼于地形简化中的不确定性对简化DEM精度的影响。地形简化中存在不确定性。一方面原始DEM存在不确定性,导致简化DEM同样存在不确定性;另一方面,简化模型的不同也会导致结果的不确定性。为探究这两种不确定性在地形简化中的影响,本文首先使用序贯高斯模拟对原始DEM的不确定性进行建模,得到若干幅具有相同空间结构的数据源,并分别使用三种基于特征的地形简化方法对其进行地形简化。通过叠加等高线,计算RMSE、骨架线保留度,构建不确定性概率图等方式对两种不确定性进行了比较和分析,结果表明,两种不确定性对简化后地形的精度均有影响。简化方法的不确定性的影响大于原始DEM的不确定性的影响,特别在地形复杂的山地区域。与其他精度影响因子相比,地形地貌是简化DEM精度的最大影响因素。采样间隔是重要影响因素。简化方法的不确定性的影响有时甚至大于采样间隔的影响。插值方法的影响最小,与原始DEM的不确定性相当。
[Abstract]:Digital elevation model is not only an important spatial data, but also the main data source of terrain analysis. Scale effect results in different topographic characteristics of the same terrain at different scales, and the expression of terrain is not the same. In the multi-scale terrain modeling, the rough digital elevation model comes from the information reduction and terrain reconstruction of the fine scale digital elevation model. This process is called terrain simplification. There are many factors that affect the precision of digital elevation model, such as terrain and geomorphology, sampling interval. Interpolation method and so on. It is very important to know the influence of these factors in the process of building digital elevation model or terrain analysis based on digital elevation model. It is still necessary to consider the influence of these factors. Some scholars have studied such factors as topography, sampling interval, interpolation method and spatial resolution. This paper focuses on the influence of the uncertainty in terrain simplification on the precision of simplified DEM. There are uncertainties in terrain simplification. On the one hand, there are uncertainties in the original DEM. There is also uncertainty in simplifying DEM; On the other hand, the difference of simplified model will lead to uncertainty of the result. In this paper, we first use sequential Gao Si simulation to model the uncertainty of the original DEM, and obtain some data sources with the same spatial structure. Using three feature-based terrain simplification methods to simplify the terrain, the RMSE, skeleton line retention is calculated by superposition contour. The two kinds of uncertainty are compared and analyzed by constructing uncertainty probability map, and the results show that. Both uncertainties have an effect on the precision of the simplified terrain. The uncertainty of the simplified method is greater than that of the original DEM. Especially in mountainous areas with complex terrain. Compared with other precision factors. Topography and geomorphology are the most important factors to simplify the precision of DEM. Sampling interval is an important factor. The influence of uncertainty of simplification method is sometimes greater than that of sampling interval. The influence of interpolation method is the least. The uncertainty is comparable to that of the original DEM.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P208

【相似文献】