遥感图像多维量化关联规则挖掘

数据与数据库的爆炸式增长引发了一个十分突出的问题,就是如何高效、智能地将海量的数据转化为有用的信息和知识?近年来,数据挖掘技术的广泛研究正是基于这个目的。初步研究了卫星遥感数据的关联规则挖掘及其在土壤侵蚀和退耕还林上的应用。根据多维空间数据的特点,将遥感数据的属性值划分为不同的块。同时为了充分利用现有的关联规则挖掘的算法,还将划分好的数据转变为事务数据库形式。最后,利用Apriori算法提取了土壤侵蚀强度与坡度、植被覆盖度以及坡耕地之间有意义的关联,为退耕还林还草决策提供有益的支持。

Multidimensional and Quantitative Association Rule Mining from Remote Sensing Images

Data mining and knowledge discovery from a large amount of image data such as remote sensing images has become highly required recent years. The association rule discovery problem in particular has been widely studied. This paper presents preliminary work in using data mining techniques to find interesting multidimensional and quantitative association rule from remotely sensed data. The data concludes two variables (e.g. vegetation cover and farmland distribution) extracted from Landsat 7 ETM, slope (from DEM data) and along with additional data from USLE model (e.g. soil erosion ). Based on the characteristics of the remote sensed data, this article presents a method partitioning quantitative attributes into unequal partitions. We show one way to generate association rule is to transform the images' data into a set of market\|basket type transaction. The main advantage of doing this is that we can use the existing algorithm and software to discover the association rule that exist in the data. One example about soil and water erosion and closing farming land for forest or grasses is employed to support it.