资源一号02C星数据在土地利用分类中的应用  

Application of ZY-1 02C Satellite Imagery in Land Use Classification

作　　者：印影[1] 姜琦刚[1] 林楠[1] 崔璨[1] 

机构地区：[1]吉林大学综合信息矿产预测研究所和地球探测科学与技术学院,长春130026

出　　处：《科学技术与工程》2014年第29期260-264,共5页Science Technology and Engineering

摘　　要：以资源一号02C星的2.36mHRC和10mMUX影像为数据源,根据试验区特点进行数据预处理,得到同时具有多光谱和高分辨率特征的影像,采用面向对象的分类方法进行土地利用分类,总结资源一号02C星的数据特点及应用潜力。结果显示:Pansharp融合法能有效保持02C影像的空间和光谱信息;面向对象的分类方法能充分利用02C高分辨率融合影像具有的丰富光谱、形状、纹理等特征信息,总体分类精度达到93.04%,Kappa系数达到0.89,结果优于传统基于像元的分类法。表明资源一号02C影像能较好地应用于土地利用宏观监测。ZY-102C 2.36m high-resolution and 10m multi-spectral imagery were chosen as data sources.According to the characteristics of the experimental zone,the images were processed to get the multi-spectral and high-resolution images.Then the land use information was extracted by object-oriented classification.The features and the applied potential of ZY-102C data were studied at last.Results show that：the optimal fusion method is Pansharp; object-oriented classification method can make full use of ZY-102C high resolution imagery,the classification accuracy and Kappa coefficient are 93.04% and 0.89,better than the traditional supervised classification.These all prove that the ZY-102C data can effectively apply to the land-use macroscopic monitoring.

关 键 词：资源一号02C 高分辨率 遥感 土地利用 面向对象 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]