论文部分内容阅读
精确的土地覆被分类结果是研究煤田火区生态环境变化的基础。本文基于Landsat8卫星遥感数据,依据地形特征、主体地物类型以及辐射特征,将乌达煤田分为五个子区。利用多光谱特征、高程、坡度和热辐射特性构建决策树模型,并分区实现土地覆被分类。分类结果表明,与传统决策树分类法相比,基于决策树法的分区分类方法总体分类精度提高了14.75%,Kappa系数增加了0.17,其准确性有了较大的提高。
The accurate land cover classification results are the basis for studying the changes of ecological environment in coalfield fire area. Based on the Landsat8 satellite remote sensing data, the paper divides the Wuda coalfield into five sub-areas based on the topographic features, the types of the main objects and the radiation characteristics. The multi-spectral features, elevation, slope and heat radiation characteristics are used to construct the decision tree model, and the classification of land cover is implemented. The classification results show that compared with the traditional decision tree classification method, the classification accuracy of the classification method based on the decision tree method is increased by 14.75% and the Kappa coefficient is increased by 0.17, and its accuracy is greatly improved.