基于随机森林的洪河湿地遥感影像分类研究

摘要

Random Forests is one of the most effective methods of classification .It attracts researchers from different backgrounds and has been widely applied to many disciplines .A Random Forest ( RF) classifier was applied to spectral extracted from Landsat TM im-agery to increase the accuracy of Honghe wetland image classification .The result of RF is compared with the supervised classification techniques including maximum likelihood classification (MLC) and classification and regression tree (CART).This research indi-cates that RF performs relatively better than MLC and CART , providing overall accuracy of 88.31% and kappa values of 0.82.RF can improve the classification accuracy of remote sensing images and can be applied in the study of wetland information extraction .%随机森林（ Random Forests ）是一种最有效的分类方法之一。现阶段，它吸引了来自不同领域的研究人员，被广泛应用到不同的学科领域之中。本文采用TM影像，运用随机森林算法，对洪河湿地影像进行分类，并与最大似然监督分类方法（ Maximum Likelihood Classification ，MLC）和 CART （ Classification And Regression Tree ）算法对比。结果表明，基于RF算法的分类结果的总精度和Kappa系数分别为88．31％和0．82，较MLC和CART分类方法有明显提高。从而证明RF算法可以提高遥感影像的分类精度，并可应用在湿地信息的提取研究中。