Please wait a minute...
img

官方微信

遥感技术与应用  2014, Vol. 29 Issue (2): 315-323    DOI: 10.11873j.issn.1004-0323.2014.2.0315
图像与数据处理     
结合纹理特征的SVM样本分层土地覆盖分类
刘萌萌1,刘亚岚1,孙国庆2,3,彭立2,3
(1.中国科学院遥感与数字地球研究所 国家环境保护卫星遥感重点实验室,北京 100101;
2.中南大学,湖南 长沙 410083;3.湖南省交通规划勘察设计院,湖南 长沙 410008)
SVM Land Cover Classification based on Spectral and Textural Features Using Stratified Samples
Liu Mengmeng1,Liu Yalan1,Sun Guoqing2,3,Peng Li2,3
(1.Institute of Remote Sensing and Digital Center,Chinese Academy of Sciences,Beijing 100101,China;
2.Central South University,Changsha 410083,China;
3.Hunan Provincial Communications Planning Survey & Design Institute,Changsha 410008  China)
 全文: PDF(3041 KB)  
摘要:

支持向量机(SVM)分类在精度、泛化性、高维数据处理等方面都具有较强的优势,在遥感影像分类中也得到了广泛应用。由于遥感影像“同物异谱”和“异物同谱”现象的影响,结合纹理特征提高SVM分类精度已成为遥感应用研究的热点。但不同尺度的纹理特征突出的信息不一,在同一尺度上难以区分的地物在多尺度空间则更容易区分,因此,采用多尺度纹理特征进行SVM分类,并从分类样本和纹理特征的选取两个方面探讨SVM土地覆盖分类的方法。首先,以ALOS影像为例,通过灰度共生矩阵提取不同尺度、不同方向的几种纹理特征;然后在光谱分类结果基础上,借助地类特征曲线,选取有效的多尺度纹理特征,最后进行样本分层分类。样本分层分类是选取首层样本进行分类,再从“漏分和错分”地块中选取新样本加入到首层样本中,得到第二层样本并对整个影像进行分类;用同样的方法选出第三层样本或更高层样本进行分类,直到结果满意为止。结果表明:该方法比仅用光谱特征的SVM分类总精度提高了8.11%,Kappa系数增加了0.11。其中,纹理特征的引入使分类总精度提高了4.13%,且对纹理特征较明显的地类更有效;采用样本分层后的分类总精度进一步提高了3.98%,且各单一地类的精度也都有不同程度的提高。借助地类特征曲线选择合适的纹理特征具有一定的可行性,并且采用样本分层的方法能够提高SVM分类的精度。

关键词: 纹理特征SVM样本分层遥感影像分类多尺度    
Abstract:

Support vector machine (SVM) shows great performance in many classification algorithms,with the merits of high precision,generalization ability and high\|dimensional data processing ability.Therefore,It has been widely used in remote sensing classifications.SVM classification,combining with texture features,has been the research focus of remote sensing applications.Since texture features can overcome the phenomena of “the same thing with different spectrum and different things with the same spectrum” in remote sensing images.Multi\|scale texture features were used to distinguish features in different scales space,which were difficult to distinguish in single scale.The study was mainly focus on texture features selection and classification with stratified samples.Firstly,using ALOS pan and multispectral remote sensing images,8 kinds of texture features in different scales and directions were extracted,based on the Gray Level Co\|occurrence Matrix;Secondly,with the help of the characteristic curve of land types,texture features of mean,homogeneity and dissimilarity in multi\|scale were selected,based on the spectral classification results.Finally,the sample stratification method was used in the SVM classification of land cover,which combined spectral with these three kinds of multi\|scale texture features in different directions.The sample stratification was implemented as follows:firstly,select the initial training samples to classify the land types;secondly,select new samples from the misclassified plots of the initial classification results,then put these new samples and the initial training samples together constituted the second-level training samples.If the second\|level training samples met the needs of classification,they were the final training samples.If not,using the same method to select higher-level training samples.The experimental results show that the overall accuracy and Kappa coefficient of the SVM classification,which combined spectral with multi-scale texture features,using the third\|level training samples,which are improved by 8.11% and 0.11 respectively,compared with that based only on spectral features,using the initial level training samples.Texture features made overall classification accuracy improved 4.13%,especially for those land cover types whose texture features is strong;while stratified sample contributed to 3.98%,and the accuracy of single classification has different degrees of improvement.Results illustrates the method used in this paper is effective.

Key words: Texture features    SVM    Stratified samples    Remote sensing images classification    Multi-scale
收稿日期: 2013-03-02 出版日期: 2014-05-14
:  TP 753  
基金资助:

国家自然科学基金项目(40971201)。

通讯作者: 刘亚岚(1968-),女,湖南常德人,研究员,博士,主要从事遥感图像信息提取与图像理解及其应用。Email:liuyl@irsa.ac.cn。   
作者简介: 刘萌萌(1989-),女,河南开封人,硕士研究生,主要从事遥感图像分类和信息提取研究。Email:liumm@irsa.ac.cn。
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  
刘萌萌
刘亚岚
孙国庆
彭立

引用本文:

刘萌萌,刘亚岚,孙国庆,彭立. 结合纹理特征的SVM样本分层土地覆盖分类[J]. 遥感技术与应用, 2014, 29(2): 315-323.

Liu Mengmeng,Liu Yalan,Sun Guoqing,Peng Li. SVM Land Cover Classification based on Spectral and Textural Features Using Stratified Samples. Remote Sensing Technology and Application, 2014, 29(2): 315-323.

链接本文:

http://www.rsta.ac.cn/CN/10.11873j.issn.1004-0323.2014.2.0315        http://www.rsta.ac.cn/CN/Y2014/V29/I2/315

[1] Hu Yufu,Deng Liangji,Kuang Xianhui,et al.Study on Land Use Classification of High Resolution Remote Sensing Image based on Texture Feature[J].Geography and Geo-information Science,2011,27(5):42-45.[胡玉福,邓良基,匡先辉,等.基于纹理特征的高分辨率遥感图像土地利用分类研究[J].地理与地理信息科学,2011,27(5):42-45.]

[2]Lin Xiaofeng.Classification Algorithms based on Texture Features for Remote Sensing Image[D].Dalian:Dalian University of  Technology,2007.[林晓峰.基于纹理特征的遥感图像分类算法研究[D].大连:大连理工大学,2007.]

[3]Huang Xin,Zhang Liangpei,Li Pingxiang.Classification of High Spatial Resolution Remotely Sensed Imagery based Upon Fusion of Multiscale Features and SVM[J].Journal of Remote Sensing,2007,11(1):48-54.[黄昕,张良培,李平湘.基于多尺度特征融合和支持向量机的高分辨率遥感影像分类[J].遥感学报,2007,11(1):48-54.][4]Li Xiaowen,Cao Chunxiang,Zhang Hao.Research Advance in the Scale Issue[J].Journal of Remote Sensing,2009,(Sup.):12-20.[李小文,曹春香,张颢.尺度问题研究进展[J].遥感学报,2009,(增刊):12-20.]

[5]Chen Chen,Zhang Youjing.The Classification of SVM based on the Multi-scale Texture Information and Spectral Information[J].Science of Surveying and Mapping,2009,34(1):29-31.[陈晨,张友静.基于多尺度纹理和光谱信息的SVM分类研究[J].测绘科学,2009,34(1):29-31.]

[6]Zhang Jinshui,He Chunyang,Pan Yaozhong,et al.The High Spatial Resolution RS Image Classification based on SVM Method with the Multi-source Data[J].Journal of Remote Sensing,2006,10(1):49-57.[张锦水,何春阳,潘耀忠,等.基于SVM的多源信息复合的高空间分辨率遥感数据分类研究[J].遥感学报,2006,10(1):49-57.]

[7]Huang Xin,Zhang Liangpei,Li Pingxiang.Classification of High Spatial Resolution Remotely Sensed Imagery based on the Fusion of Spectral and Shape Features[J].Journal of Remote Sensing,2007,11(2):193-200.[黄昕,张良培,李平湘.融合形状和光谱的高空间分辨率遥感影像分类[J].遥感学报,2007,11(2):193-200.][8]Liu Feng,Gong Jianya.A Classification Method for High Spatial Resolution Remotely Sensed Image based on Multi-feature[J].Geography and Geo-information Science,2009,25(3):19-22.[刘峰,龚健雅.一种基于多特征的高光谱遥感图像分类方法[J].地理与地理信息科学,2009,25(3):19-22.]

[9]Tang Yinfeng,Huang Zhiming,Huang Rongjuan,et al.Texture Image Classification based on Multi-feature Extration and SVM Classifier[J].Computer Application and Software,2011,28(6):22-25.[唐银凤,黄志明,黄荣娟,等.基于多特征提取和SVM分类器的纹理图像分类[J].计算机应用与软件,2011,28(6):22-25.]

[10]Long Haixiang,Gao Xin.SVM Classification of SAR Image based on Texture and Internal Edge[J].Application Research of Computers,2011,28(9):3551-3553.[龙海翔,高鑫.基于纹理和边缘的SAR图像SVM分类[J].计算机应用研究,2011,28(9):3551-3553.]

[11]Song X F,Duan Z,Jiang X G.Comparison of Artificial Neural Networks and Support Vector Machine Classifiers for Land Cover Classification in Northern China Using a SPOT-5 HRG Image[J].International Journal of Remote Sensing,2012,33(10):3301-3320.

[12]Zhang Rui,Ma Jianwen.State of the Art on Remotely Sensed Data Classification based on Support Vector Machines[J].Advances in Earth Science,2009,24(5):555-562.[张睿,马建文.支持向量机在遥感数据分类中的应用新进展[J].地球科学进展,2009,24(5):555-562.]

[13]Huang C,Davis L S,Townshend J R G.An Assessment of Support Vector Machines for Land Cover Classification[J].International Journal of Remote Sensing,2002,23(4):725-749.

[14]Giles M F,Ajay M.Toward Intelligent Training of Supervised Image Classifications Directing Training Data Acquisition for SVM Classification[J].Remote Sensing of Environment,2004,93:107-117.

[15]Ren Y H,Liu Y L,Fan J C,et al.A Study on Land Cover Classification based on HJ-1 CCD Image[C]//International Geoscience and Remote Sensing Symposium (IGARSS),2009,3:III408-III411.

[16]Wang Zuocheng.Research on Classification of Remote Sensing Image with Texture[D].Chengdu:Southwest Jiaotong University,2007.[王佐成.基于纹理的遥感图像分类研究[D].成都:西南交通大学,2007.]

[17]Lin Shengliang,Liu Zhi.Parameter Selection in SVM with RBF Kernal Function[J].Journal of Zhejiang University of Technology,2007,35(2):163-167.[林升梁,刘志.基于RBF核函数的支持向量机参数选择[J].浙江工业大学学报,2007,35(2):163-167.]

[18]Keuchel J,Naumann S,Heiler M,et al.Automatic Land Cover Analysis for Tenerife by Supervised Classification Using Remotely Sensed Data[J].Remote Sensing of Environment,2003,86:530-541.

[19]Su L H,Chopping M J,Rango A,et al.Support Vector Machines for Recognition of Semi-arid Vegetation Types Using MISR Multi-angle Imagery[J].Remote Sensing of Environment,2007,107:299-311.

[20]Oommen T,Misra D,Twarakavi N K C,et al.An Objective Analysis of Support Vector Machine based Classification for Remote Sensing[J].Mathematical Geosciences,2008,40:409-424.

[21]Zhang Miao,Jiang Zhirong,Ma Mingguo,et al.Fine Classification of Planting Structure in the Middle Reaches of Heihe River Basin based on Hyperspectral Compact Airborne Spectrographic Imager(CASI)Data[J].Remote Sensing Technology and Application,2013,28(2):283-289.[张苗,蒋志荣,马明国,等.基于CASI影像的黑河中游种植结构精细分类研究[J].遥感技术与应用,2013,28(2):283-289.]

[22]Zhao Yingshi.The Principle and Method of Analysis of Remote Sensing Application[M].Beijing:Science Press,2003:207-208.[赵英时.遥感应用分析原理与方法[M].北京:科学出版,2003:207-208.]

[23]Jensen J R.Introductory Digital Image Processing:A Remote Sensing Perspective(Edition 3)[M].Beijing:Science Press,2007:506-508.


 


 

[1] 任浙豪,周坚华. 增大特征空间复杂度的方法——以城镇下垫面遥感分类为[J]. 遥感技术与应用, 2018, 33(3): 408-417.
[2] 王宁,陈方,于博. 基于形态学开运算的面向对象滑坡提取方法研究[J]. 遥感技术与应用, 2018, 33(3): 520-529.
[3] 何艺,周小成,黄洪宇,许雪琴. 基于无人机遥感的亚热带森林林分株数提取[J]. 遥感技术与应用, 2018, 33(1): 168-176.
[4] 姜萍,刘修国,陈启浩,邵芳芳. 利用多尺度SVM-CRF模型的极化SAR图像建筑物提取[J]. 遥感技术与应用, 2017, 32(3): 475-482.
[5] 朱济帅,尹作霞,谭琨,王雪,李二珠,杜培军. 基于空间邻域信息的高光谱遥感影像半监督协同训练[J]. 遥感技术与应用, 2016, 31(6): 1122-1130.
[6] 翟玮,沈焕锋,黄春林. 结合PolSAR影像纹理特征分析提取倒塌建筑物[J]. 遥感技术与应用, 2016, 31(5): 975-982.
[7] 邓滢,张红,王超,刘萌. 结合纹理与极化分解的面向对象极化SAR水体提取方法[J]. 遥感技术与应用, 2016, 31(4): 714-723.
[8] 赵彦博,曹学诚,南卓铜,吴小波. 基于开源GIS的DHSVM模型河网数据自动制备方法应用研究[J]. 遥感技术与应用, 2016, 31(4): 793-800.
[9] 林志垒,晏路明. 高光谱影像的BDT-SVM地物分类算法与应用[J]. 遥感技术与应用, 2016, 31(1): 177-185.
[10] 张文博,覃志豪,刘含海. 基于分形理论的土地利用类型变化研究—以广州番禺为例[J]. 遥感技术与应用, 2015, 30(5): 952-958.
[11] 费鲜芸,王婷,魏雪丽. 基于多尺度分割的遥感影像滨海湿地分类[J]. 遥感技术与应用, 2015, 30(2): 298-303.
[12] 宋晓阳,姜小三,江东,黄耀欢,万华伟,王昌佐. 基于面向对象的高分影像分类研究[J]. 遥感技术与应用, 2015, 30(1): 99-105.
[13] 邹亚荣,黄磊,张治平. 结合纹理特征的SVM岛礁信息提取分析[J]. 遥感技术与应用, 2014, 29(5): 812-817.
[14] 刘朝相,宫兆宁,赵文吉. 基于SVM模型的妫水河叶绿素a浓度的遥感反演[J]. 遥感技术与应用, 2014, 29(3): 419-427.
[15] 许夙晖,慕晓冬,柯冰,王晓日. 基于遥感影像的军事阵地动态监测技术研究[J]. 遥感技术与应用, 2014, 29(3): 511-516.