基于数据挖掘的遥感影像围填海智能检测方法研究

以Landsat TM遥感影像为数据源,用本文提出的方法对渤海区域进行了检测。首先根据初始时刻影像（例如2001年影像）检测出海水区域,即感兴趣区域,然后以终止时刻影像（例如2007年影像）的同一感兴趣区域为研究对象,检测其中的陆地区域,即围填海区域。为了能够检测出围填海区域,采用数据挖掘中的关联规则算法,挖掘了围填海检测规则。为了验证本文提出的围填海区域检测方法的有效性,以整个渤海为研究区域,采用挖掘出的检测规则对待检测影像进行围填海区域检测,检测结果表明本文提出的检测方法能较高精度的检测出围填海区域。

The Sea-filling Land Detect Method Research Using Remote Sensing Images Based on Data Mining

By using remote sensing images of Landsat TM and the method proposed in this paper,the Bohai sea area is tested.First,the sea area which is called interested region is detected according to the original image(such as image of 2001 year),and then,the land area which is called sea-filling land area is detected on the interested region which gained from the original image according to the final image(such as image of 2007 year).In order to be able to detect the sea-filling land area,the association rules algorithm is used to mine the detect rules of sea-filling land area.The validity of this method proposed in this paper is confirmed by using,the mining rules with the whole Bohai sea area to detect the sea-filling land area.The result shows that the method can achieve high accuracy of the test to detect the sea-filling land area.