期刊文献+

一种基于谱域-空域组合特征支持向量机的高光谱图像分类算法 被引量:13

Hyperspectral Image Classification Algorithm Based on Spectral-Spatial Hybrid Features and SVM
下载PDF
导出
摘要 针对高光谱图像分类问题,提出了一种基于支持向量机的利用组合特征对高光谱图像进行分类的算法,组合特征综合了高光谱图像的光谱域和空域信息。针对图像的高维数据特性,利用最大噪声分量方法进行特征提取,对得到的主分量图像,采用虚拟维数估计算法来确定需要保留的主分量数目,并用数学形态学操作用保留的主分量图像中提取目标的形态信息。根据得到的空域特征并结合原始光谱域特征,构造用于分类的组合特征矢量。而且采用了支持向量机,利用了其支持小样本、效率高的优点。高光谱数据实验表明,本文提出的方法和单独使用谱域或空域信息进行分类相比,具有一定的优越性。 A hyperspectral image classification algorithm based on support vector machine(SVM) is presented.The hybrid features are made up of the spectral domain and spatial domain information of the image.In order to deal with high dimensional data of the hyperspectral image,a maximum noise fraction(MNF) extraction method is adopted.A virtual dimensionality(VD) estimation algorithm is used to determine the number of the principal components(PC) remained.The morphology profiles(MP) of the object are extracted from the remained PC images by using mathematical morphology method.The hybrid features for image classification are constructed by combining the spatial characteristics with spectral domain characteristics of the original image.Because of its high efficiency and small training samples,the SVM is adopted as the classifier.The superiors of the proposed method compared with the classifiers only using spectral or spatial information are shown by experiments.
出处 《宇航学报》 EI CAS CSCD 北大核心 2011年第4期917-921,共5页 Journal of Astronautics
基金 国家自然科学基金(40901216) 湖南省研究生科研创新项目(CX2010B020) 国防科技大学博士研究生创新基金(B100402)
关键词 组合特征 数学形态学 最大噪声分量 虚拟维数 Hybrid features Mathematical morphology Maximum noise fraction Virtual dimensionality
  • 相关文献

参考文献11

  • 1谌德荣,陶鹏,张立燕,范宁军.基于区域特征光谱的自组织特征映射神经网络高光谱图像主成分提取方法[J].宇航学报,2007,28(6):1689-1692. 被引量:3
  • 2Mathieu F,Jón A B,Jocelyn C,et al.Spectral and spatial clas-sification of hyperspectral data using SVMs and morphologicalprofiles. IEEE Transactions on Geoscience and Remote Sens-ing . 2008
  • 3Santiago V F,,Vidya M.Improving hyperspectral image classifi-cation using spatial preprocessing. IEEE Geoscience and Re-mote Sensing Letters . 2009
  • 4Devis T,Fabio P,Mikhail K,et al.Classification of very highspatial resolution imagery using mathematical morphology andsupport vector machines. IEEE Transactions on Geoscienceand Remote Sensing . 2009
  • 5GonzalezR C,WoodsR E.Digital image processing. . 2008
  • 6ftp://ftp.ecn.purdue.edu/biehl/MultiSpec/92AV3C/[OL] .
  • 7Gustavo C V,Luis G C,Jordi M M,Joan V F,Javier C M.Composite kernels for hyperspectral image classifica-tion. IEEE Transactions on Geoscience and Remote Sens-ing Letters . 2006
  • 8Chang, C.I,Du, Q.Estimation of Number of Spectrally Distinct Signal Sources in Hyperspectral Imagery. IEEE Transactions on Geoscience and Remote Sensing . 2004
  • 9Melgani F,Bruzzone L.Classification of Hyperspectral Remote Sensing Images with Support Vector Machines. IEEE Transactions on Geoscience and Remote Sensing . 2004
  • 10Green AA,Berman M,Switzer P,et al.A transformation for ordering multispectral data in terms of image quality with implications for noise removal. IEEE Transactions on Geoscience and Remote Sensing . 1988

二级参考文献2

  • 1[2]Luis O,Jiménez,Jorge L,Rivera-Medina,Eladio Rodríguez-Diaz,et al.Integration of spatial and spectral information by measns of unsupervised extraction and classification for homogenous objects applied to multispectral and hyperspectral data[J].IEEE Transactions on Geoscience and Remote Sensing,2005,43(4):844-851
  • 2[6]KohonenT.The self-Organiziug Map.ProcIEEE,1990,78:1464-14801

共引文献2

同被引文献151

引证文献13

二级引证文献66

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部