多特征组合的TM影像EnMap-Box土地利用分类
|
摘要
为研究多特征组合对提高遥感影像土地利用分类精度的作用,以云南省洱源县作为研究区域,利用EnMAP-Box软件对选取的多特征组合向量进行支持向量机(support vector machines,SVM)分类。本文选取了绿度植被指数、归一化建筑指数及基于灰度共生矩阵提取的纹理信息和最优波段组合等光谱特征构成分类多特征组合向量,通过EnMAP-Box软件寻优SVM最佳分类模型对多特征组合向量进行遥感影像土地利用分类。同时选择了云南省思茅区验证此法的适用性。结果表明,基于多特征组合的支持向量机分类法其总体分类精度为90.73%,分别比最大似然分类法高13%左右,比原始波段影像的分类精度高大约7%左右,另一验证区域精度结果表明此法具有一定适用性。
In order to research the impact of multi-feature selection on remote sensing classification with land use, the Eryuan inYunnan Province is selected as the research area. EnMAP-Box is used to classify the selected multi-feature combination vectors. In this research, the greenness vegetation index, the normalized building index and texture features and optimum band spectral features are selected to make up classification multi-feature vectors. The best SVM classification model is optimized by EnMAP-Box software to classify the land use RS images. The results show that the overall classification accuracy of support vector machine classification based on multi-feature combination is 90.73%, which is about 13% higher than the maximum likelihood method and 7% higher than the original band image. And another verification area accuracy result indicates that this method has certain universal applicability.
In order to research the impact of multi-feature selection on remote sensing classification with land use, the Eryuan inYunnan Province is selected as the research area. EnMAP-Box is used to classify the selected multi-feature combination vectors. In this research, the greenness vegetation index, the normalized building index and texture features and optimum band spectral features are selected to make up classification multi-feature vectors. The best SVM classification model is optimized by EnMAP-Box software to classify the land use RS images. The results show that the overall classification accuracy of support vector machine classification based on multi-feature combination is 90.73%, which is about 13% higher than the maximum likelihood method and 7% higher than the original band image. And another verification area accuracy result indicates that this method has certain universal applicability.
引文
[1] Townshend J R,Masek J G,Huang C,et al.Global Characterization and Monitoring of Forestcover Using Landsat Data;Opportunnies and Challenges[J].International Journal of Digital Earth,2012,5(5):373-397
[2] 韩立群.人工神经网络[M].北京:北京邮电大学出版社,2006
[3] Richard O,Duda P E,Hart D G.Pattern Classification,Second Edition[M].Beijing:China Machine Perss,2004
[4] 万意,李长春,赵旭辉,等.基于SVM的光学遥感影像分类与评价[J].测绘地理信息,2018,43(6):74-77
[5] 陈百明,周小萍.《土地利用现状分类》国家标准的解读[J].自然资源学报,2007,22(6):994-1 003
[6] 赵庆展,刘伟,尹小君,等.基于无人机多光谱影像特征的最佳波段组合研究[J].农业机械学报,2016,47(3):242-248
[7] Chavez P S,Berlin G L,Sowers L B.Statistical Method for Selecting Landsat MSS Ratios [J].Journal of Applied Photographic Engineering,1982,8 (1):22-30
[8] 赵英时.遥感应用分析原理与方法[M].北京:科学出版社,2013
[9] 查勇,倪绍祥,杨山.一种利用TM 图像自动提取城镇用地信息的有效方法[ J] .遥感学报,2003,7(1):37-40
[10] ZHA Yong,GAO J,Ni S.Use of Normalized Difference Built-up Index in Automatically Mapping Urban Areas from TM Imagery[J].International Journal of Remote Sensing,2003,24(3):583-594
[11] 韦玉春,汤国安,汪闽,等.遥感数字图像处理教程第二版[M].北京:科学出版社,2014
[12] 陈元鹏,郧文聚,周旭,等.基于MESMA和RF的山丘区土地利用信息分类提取[J].农业机械学报,2017,48(7):136-144
[13] Barald A,Parminggian F.An Investigation on the Texture Characteristics Associated with Gray Level Co-occurrence Matrix Statistical Parameters [J].IEEE Trans on Geoscience and Remote Sensing,1995,32(2):293-303
[14] Sebastian V D L,Rabe A,Held M,et al.The EnMAP-Box—A Toolbox and Application Programming Interface for En-MAP Data Processing[J].Remote Sensing,2015,7(9):112-119
[15] 林海晏,岳彩荣,吴晓晖,等.基于EnMAP-Box的遥感图像分类研究[J].西南林业大学学报,2014(2):67-71
[2] 韩立群.人工神经网络[M].北京:北京邮电大学出版社,2006
[3] Richard O,Duda P E,Hart D G.Pattern Classification,Second Edition[M].Beijing:China Machine Perss,2004
[4] 万意,李长春,赵旭辉,等.基于SVM的光学遥感影像分类与评价[J].测绘地理信息,2018,43(6):74-77
[5] 陈百明,周小萍.《土地利用现状分类》国家标准的解读[J].自然资源学报,2007,22(6):994-1 003
[6] 赵庆展,刘伟,尹小君,等.基于无人机多光谱影像特征的最佳波段组合研究[J].农业机械学报,2016,47(3):242-248
[7] Chavez P S,Berlin G L,Sowers L B.Statistical Method for Selecting Landsat MSS Ratios [J].Journal of Applied Photographic Engineering,1982,8 (1):22-30
[8] 赵英时.遥感应用分析原理与方法[M].北京:科学出版社,2013
[9] 查勇,倪绍祥,杨山.一种利用TM 图像自动提取城镇用地信息的有效方法[ J] .遥感学报,2003,7(1):37-40
[10] ZHA Yong,GAO J,Ni S.Use of Normalized Difference Built-up Index in Automatically Mapping Urban Areas from TM Imagery[J].International Journal of Remote Sensing,2003,24(3):583-594
[11] 韦玉春,汤国安,汪闽,等.遥感数字图像处理教程第二版[M].北京:科学出版社,2014
[12] 陈元鹏,郧文聚,周旭,等.基于MESMA和RF的山丘区土地利用信息分类提取[J].农业机械学报,2017,48(7):136-144
[13] Barald A,Parminggian F.An Investigation on the Texture Characteristics Associated with Gray Level Co-occurrence Matrix Statistical Parameters [J].IEEE Trans on Geoscience and Remote Sensing,1995,32(2):293-303
[14] Sebastian V D L,Rabe A,Held M,et al.The EnMAP-Box—A Toolbox and Application Programming Interface for En-MAP Data Processing[J].Remote Sensing,2015,7(9):112-119
[15] 林海晏,岳彩荣,吴晓晖,等.基于EnMAP-Box的遥感图像分类研究[J].西南林业大学学报,2014(2):67-71