中文摘要：

针对利用Morton编码规则压缩数字高程模型（DEM）数据存在的只适用于较平缓地形的问题，提出了基于地表粗糙度的Morton分块压缩方法．在地形分块时采用二级划分方法，只对地表粗糙度比较小的地形块编码压缩，减少数据读取及编码数量，提高压缩速率；在建模部分，首先对基于地形局部熵的生成准则进行改进，形成一种视相关局部熵生成准则，利用此准则对近处地形进行简化，然后将此准则与Phong光照模型相结合形成一种混合生成准则，用Phong光照模型对远处地形简化．实验结果表明，采用文中方法可以在有效地减少绘制三角形的同时保留地形局部细节，提高了地形可视化效果．

英文摘要：

The method of compressing digital elevation model （DEM） using Morton code can only be applied to the relatively flat terrain. In order to avoid the limitation of the method, this paper presents a Morton compressing method in terrain data block partition based on the surface roughness. The 2- level dividing method is used in terrain partition. Only the terrain blocks with lower surface roughness are encoded and compressed, which could reduce the amount of reading DEM data and improve the compression speed. In the process of modeling, the generating rule based on the local terrain entropy is discussed and improved to form a view-dependent local entropy rule. Then, combining the improved rule with Phong illumination model into our hybrid rule, the algorithm in this paper uses viewdependent local entropy rule to simplify terrain near and Phong illumination model to simplify terrain far. The experimental results demonstrate that the proposed method can keep the local details of terrain and simultaneously reduce rendering triangles efficiently, improving effects of terrain visualization.