基于高分一号影像的雄安新区土地利用分析

Analysis on Land Use in Xiongan New Area Based on GF-1 Satellite

土地利用的分类精度对土地资源开发有很大影响。选取雄安新区2017年5月12日的高分一号影像作为试验数据,分别运用监督分类方法中的最大似然分类器和面向对象的分类方法对影像进行土地利用分类,得到研究区域的土地利用分类情况。面向对象分类方法的Kappa系数和总体分类精度都高于最大似然分类器的分类结果,分别达到了0.968 8和97.500 0%。实验结果表明:针对高分一号影像,对比最大似然分类器,使用面向对象的分类能提高研究区域的土地利用分类精度。面向对象分类结果可以为雄安新区的土地利用分类提供参考。

The accuracy of land use classification has a great impact on the development of land resources.In this paper,we select GF-1 image of Xiongan new area in May 12,2017 as the test data,and use the maximum likelihood classifier of supervised classification and object-oriented classification method respectively to do the classification of land use.The classification of land use in the study area is obtained.The Kappa coefficient and the overall classification accuracy of the object-oriented classification method are higher than the classification results of the maximum likelihood classifier,and they reach to 0.968 8 and 97.500 0%respectively.The results of the experiment show that:for GF-1 image,by comparing the maximum likelihood classifier,the method of using object-oriented classification can improve the land use classification accuracy of the study area.The results of object-oriented classification can provide reference for land use classification in Xiongan new area.