偃师市土地利用遥感分类研究  

Study on remote sensing classification of land utilization in Yanshi city

作　　者：王国重[1] 李中原 张继宇 程焕玲 李佳红 WANG Guo-zhong;LI Zhong-yuan;ZHANG Ji-yu;CHENG Huan-ling;LI Jia-hong(Yellow River Scientific Research Institute of Hydrology and Water Resources,Zhengzhou450004,China;Hydrology and Water Resources Bureau of Henan Province,Zhengzhou450003,China;Henan Yellow River Bureau,Zhengzhou450003,China;Soil and Water Conservation Supervision Central Station of Henan Province,Zhengzhou450008,China;Hydrology and Water Resources Survey Bureau of Jiaozuo,Henan Province,Jiaozuo454002,Henan,China)

机构地区：[1]黄河水文水资源科学研究院,郑州450004 [2]河南省水文水资源局,郑州450003 [3]河南黄河河务局,郑州450003 [4]河南省水土保持监测总站,郑州450008 [5]河南省焦作水文水资源勘测局,河南焦作454002

出　　处：《湖北农业科学》2021年第13期60-62,67,共4页Hubei Agricultural Sciences

基　　金：河南省科技攻关计划项目(GG201719)。

摘　　要：为了掌握河南省偃师市真实、准确的土地利用数据,根据其Landsat8影像,采用支持向量基(SVM)和最大似然法(ML)对其2016年用地类型进行分类。结果表明,2种方法都具有较高的分类精度,但SVM的分类精度略低,这可能与所选参数、核函数及训练样本的影响有关,需做进一步研究。To grasp true and accurate land data in Yanshi city,Henan province,the classing methods were adopted by support vector(SVM)and maximum likelihood(ML)to classify its land use types in 2016 with landsat8 images.The results showed that both methods had high classification accuracy,but the SVM classification accuracy was slightly lower,which could be related to the selected parameters,the influence of kernel function and the training sample.

关 键 词：支持向量基(SVM) 最大似然法 遥感影像 分类 偃师市 

分 类 号：P237[天文地球—摄影测量与遥感]