基于TM影像的喀什地区土地利用分类

The Land Use Classification of Kashi Region Based on TM Images

以喀什地区为研究区,选取2010年9景TM影像为遥感信息源,利用支持向量机法对喀什地区的土地进行分类。最终分成7个类别,并用混淆矩阵对分类结果作精度评价,总体分类精度为85.28%,Kappa系数为82.79%。结果表明,利用支持向量机分类方法对TM影像进行喀什地区土地利用分类与制图是可行的,能较真实地反映该地区植被和土地利用的基本特征。

Taking Kashi region as the study area,and selecting the data of nine TM images in 2010 as the remote sensing information sources,the land in Kashi region was classified by support vector machine method. Seven classes were sorted in final and the confusion matrix was applied to evaluate the accuracy of classification results. The overall classification accuracy was 85.28%,and the Kappa coefficient was 82.79%. The results showed that using the support vector machine method with TM image on Kashi was available,which can truly reflect the basic characteristics of the Kashi region's vegetation and land use.