不确定性引导下的地形简化方法研究

摘要

提出一种不确定性引导下的地形简化方法。该方法利用地统计方法中特有的方差统计量对简化结果的不确定性进行量化评估，并以此作为优化目标进行地形简化操作。实验部分，从高程精度、地表基本形态以及骨架保留度三方面对新方法的有效性进行验证。结果表明：与传统的特征简化方法相比，新方法在控制地形简化的不确定性的前提下很好地保持了地形简化精度以及原始地表的骨架结构。%A new terrain generalization method, under the guidance of uncertainty, was proposed. This method uti-lized variance statistic, which was calculated by geostatistics, to evaluate the uncertainty of generalization, which acted as optimization objective to boot terrain generalization. In the experiment, the new method would be verified from three points: elevation accuracy, basic information of surface, and maintance of terrain skeleton. The result shows that, compared to other traditional feature-based generalization methods, the new method can not only control the uncertainty of generalization, but also retain the elevation accuracy and the structure of terrain skeleton.