地物波谱数据仓库系统设计研究

Research on Designing of Data Warehouse System for Object Spectrum

典型地物波谱特征是遥感仪器波段选择的依据,是遥感解译的基础。我国已经积累了大量的地物波谱数据,而且即将获得更多的地物波谱数据。如何有效地管理数据?从观测波谱数据中挖掘出观测间隙的波谱数据?该文按照参数配套原则,提出用树状结构的事实表和分维表实现数据管理。基于相关作物模型和数据挖掘方法采掘观测间隙的光谱数据,并用北京顺义野外实验SE590波谱数据为实例介绍一种地物波谱数据仓库系统的设计。

The typica l object spectrum is one of the foundations on selecting bands of remote sensing s ensors,which is the basis of remote sensing interpretation.A lot of spectrums have been accumulated during the past time in China,and more data will be got in the future.How to manage them effectively?How to mine the spectrums on no observing days from these datum?In this paper,a database organized by trees s tructure which include dimension tables and fact tables is applied to manage spe ctrum datum matching with interrelated parameters.There is a data mining resear ch based on certain crop models and data mining method.A case is given in this paper to test data warehouse designing using SE590data obtained during Shun Y i field experiment.