基于遥感与GIS的土地利用分类方法研究——以河北省安新县为例
|
摘要
研究土地利用类型分类对国民经济具有重要意义。在土地利用类型分类中,常用的方法为监督分类与非监督分类,这两种方法都可以快速地对土地利用类型行分类,但针对不同的地物二者的分类精度是不同的。文章将最大似然法与ISODATA算法的分类结果与目视解译作对比,分析得出两种方法对不同地物的分类精度,确定其适用地物。分析得出最大似然法对于居民地、水田、天然草地的划分精度较高,精度分别为:88.89%、87.64%、90.96%;ISODATA算法对坑塘、水浇地的划分精度较高,精度分别可达到:95.38%和91.61%。
It is of great significance to study the classification of land use types for the national economy. In the classification of land use types, the commonly used methods are supervised classification and unsupervised classification, both of which can quickly classify land use types, but the classification accuracy is different for different ground objects. In this paper, the classification results of maximum likelihood method and ISODATA algorithm are compared with visual interpretation, and the classification accuracy of the two methods for different ground objects is analyzed, and the suitable ground objects are determined. The results show that the maximum likelihood method has high accuracy for the division of residential land, paddy field and natural grassland, and the accuracy is 88.89%, 87.64% and 90.96%, respectively. The accuracy of ISODATA algorithm for pit pond and irrigated land is high, and the accuracy can reach 95.38% and 91.61% respectively.
It is of great significance to study the classification of land use types for the national economy. In the classification of land use types, the commonly used methods are supervised classification and unsupervised classification, both of which can quickly classify land use types, but the classification accuracy is different for different ground objects. In this paper, the classification results of maximum likelihood method and ISODATA algorithm are compared with visual interpretation, and the classification accuracy of the two methods for different ground objects is analyzed, and the suitable ground objects are determined. The results show that the maximum likelihood method has high accuracy for the division of residential land, paddy field and natural grassland, and the accuracy is 88.89%, 87.64% and 90.96%, respectively. The accuracy of ISODATA algorithm for pit pond and irrigated land is high, and the accuracy can reach 95.38% and 91.61% respectively.
引文
[1]王敏,高新华,陈思宇,等.基于Landsat 8遥感影像的土地利用分类研究———以四川省红原县安曲示范区为例[J].草业科学,2015,32(05):694-701.
[2]吴非权,马海州,沙占江,等.基于决策树与监督、非监督分类方法相结合模型的遥感应用研究[J].盐湖研究,2005(04):9-13.
[3]杨鑫.浅谈遥感图像监督分类与非监督分类[J].四川地质学报,2008(03):251-254.
[4]杨毅柠,宋微.国土调查中的遥感目视解译与矢量化方法研究[J].环境保护与循环经济,2009,29(10):20-23.
[5]李建伏,赵玉成,贺怀清.基于最大似然原理的分类属性数据分层聚类算法[J].计算机应用与软件,2015,32(03):247-252+296.
[6]曾江源.ISODATA算法的原理与实现[J].科技广场,2009(07):126-127.
[2]吴非权,马海州,沙占江,等.基于决策树与监督、非监督分类方法相结合模型的遥感应用研究[J].盐湖研究,2005(04):9-13.
[3]杨鑫.浅谈遥感图像监督分类与非监督分类[J].四川地质学报,2008(03):251-254.
[4]杨毅柠,宋微.国土调查中的遥感目视解译与矢量化方法研究[J].环境保护与循环经济,2009,29(10):20-23.
[5]李建伏,赵玉成,贺怀清.基于最大似然原理的分类属性数据分层聚类算法[J].计算机应用与软件,2015,32(03):247-252+296.
[6]曾江源.ISODATA算法的原理与实现[J].科技广场,2009(07):126-127.