摘要
研究土地利用类型分类对土地规划有着重要的意义。遥感技术对土地利用类型分类提供技术性支持。随着遥感技术的不断发展,遥感影像分类方法不断涌现,不同的分类方法会有精度不一的问题,对其在土地利用类型分类中产生了一定的影响。本文以西安市未央区为研究区域,依据landsat8 OLI影像,采用最小距离分类法、马氏距离分类法、支持向量机分类法、神经网络分类法和最大似然分类法,对研究区进行土地利用类型分类,并对其分类结果进行精度比较。结果表明:总体上看,最大似然分类法的Kappa系数和整体分类精度均高于其他分类方法,其次为神经网络分类法、支持向量机分类法、马氏距离分类法、最小距离分类法。从不同土地利用类型来看,相较于其他分类方法,最大似然法仍表现出理想的分类效果。本研究可以对土地利用分类方法的选择提供理论依据,以期为今后的土地利用工作提供决策性支持。
It is important of land planning to study the classification of land use types.Remote sensing technology provides technical support for the classification of land use types.With the continuous development of remote sensing technology,remote sensing image classification methods are emerging,and different classification methods have different precision problems,which have a certain impact on the classification of land use types.This paper selects Xi’an Weiyang as the research area,and uses the Landsat 8 OLI image to classify the land use type of the study area.The classification methods selected are minimum distance classification,Mahalanobis distance classification,support vector machine classification,neural network classification and maximum likelihood classification.And the accuracy of the classification results are compared.The results show that for the study area of this paper,the accuracy of the maximum likelihood classification method is higher than other classification methods through comparison of different precision indicators.For different land use types,the maximum likelihood method still has a good classification effect.Followed by the neural network classification.Only by selecting appropriate classification methods and improving the accuracy and efficiency of land use type classification can we provide decision-making support for future land work and better carry out land work.
作者
白宇兴
BAI Yuxing(School of Earth Sciences and Resources,Chang'an University,Xi'an 710054,China)
关键词
遥感
土地利用类型分类
分类方法
精度评价
remote sensing
classification of land use types
classification method
accuracy evaluation
作者简介
白宇兴(1995-),男,甘肃金昌人,在读硕士,研究方向:3S应用。E-mail:920276184@qq.com
