基于分形理论的DEM数据内插算法研究

Study On DEM Interpolation Algorithm Based On The Fractal Theory

本文利用分形理论与分形布朗运动原理,提出了基于子分形维数的随机中点位移内插算法,并采用不同的插值比系数来有效避免插值点附近的"尖峰现象"。应用该算法对研究区内的低分辨率数字高程模型(DEM)格网数据进行了内插计算,结果显示内插精度有一定的提高,同时生成的地形起伏平稳,并且能较逼真的反映地形表面的细节。

On basis of fraetal theory and the principle of fractional Brownian motion （FBM）, this paper puts forward an interpolation algorithm based on sub-fractional dimensions. To avoid the creasing problem with traditional method （midpoint displacement algorithm）, various coefficient scales are used in interpolating process. Experiments on digital elevation model （DEM） data are conducted, which indicate that the interpolation can improve the precision of results and image the details of virtual terrain.