|
本论文得到以下基金的支持:
1) 国家自然科学基金项目:“基于虚拟地理环境的SARS 传播与控制模拟研究(40471103)”(2005 年1 月-2007 年12 月);
2) 武汉大学测绘遥感信息工程国家重点实验室开放研究基金:“分布式大型地形实时漫游关键技术研究(WKL(04)0302)”(2005/01/01-2006/12/31);
DEM 数据压缩方法研究
摘要
随着地理信息系统、遥感、虚拟地理环境等技术的发展,数字高程模型得到了广泛的应用。但是其数据量庞大,对存储和传输提出了很高的要求。数字高程模型的压缩也因此逐步成为研究的热点之一。
本文首先分析数字高程模型的数字特征,并在此基础上对基于分形变换和小波变换的压缩方案进行了深入研究。实验结果表明,DEM 数据具有很强的相关性,在小波变换下,可实现10~19 倍的无损压缩率,并具有可分级渐进传输的特性,这对于虚拟地理环境三维建模及其基于网络的分布可视化具有非常重要的现实意义。应用该研究结果,一方面可以在不损失精度情况下实现较高的压缩比,另外一方面可以满足可视化过程中快速渐变显示的需要。
在基于分形变换压缩的研究中,本文指出,虽然利用数字高程模型自相似性的分形变换具有更高的理论压缩效率,但限于搜索匹配算法的计算复杂度,分形压缩目前离实际应用还有较大的距离。
关键词:数字高程模型,小波变换,分形变换,数据压缩,可视化
RESEARCH ON DATA COMPRESSION ALGORITHM FOR DEM
ABSTRACT
Digital Elevation Model (DEM) has been widely used in the applications of Geography Information System, Remote Sensing, Virtual Geographic Environment, etc. As most DEM data requires massive storage, the efficient compression algorithm for DEM data is a very important problem, and has drawn much attention in recent years. Based on the statistical analysis of DEM data, we investigated wavelet-based and fractal-base compression schema. Experimental results show that DEM data is high redundant and correlated. Under the wavelet transformation, we achieved 10~19 times lossless compression with scalable transmit ability, which is meaningful for 3D modeling and visualizing of virtual geographic environment. First, higher compression ratio can be achieved with lossless precision. Second, this result can meet the need of quick display during visualizing. In the research of fractal-based compression, we find that fractal transformation is capable of modeling self-similarity of DEM data to achieve higher theoretical compression rate. However, due to the computational complexity of matching procedure, fractal- based compression is now less practical in real applications.
KEY WORDS: digital elevation model, wavelet, fractal, compression, visualization
|