高光谱遥感影像支持向量机分类中噪声影响的研究Impacts of noises on the performance of support vector machine classifier for hyperspectral remote sensing image
王小美,逄云峰,杜培军,谭琨,李光丽
摘要(Abstract)：
为了验证噪声对支持向量机分类器性能的影响,对"SVM可以有效用于含噪声和不确定性数据"这一观点进行定量分析评价,采用国产OMISII传感器获得的高光谱遥感数据进行了试验,为了更好地比较SVM分类器的抗噪性,先对原始数据进行支持向量机分类,然后在高光谱遥感影像中人为添加不同比例的椒盐噪声和条带噪声,然后进行支持向量机分类,并与传统的光谱角制图和最小距离分类算法进行比较。结果表明支持向量机具有明显的抗噪性,其分类性能受噪声影响较小,是一种有效的高光谱遥感影像分类器。
关键词(KeyWords)： 高光谱遥感;支持向量机;分类;噪声
基金项目(Foundation): 国家自然科学基金项目(40871195);; 教育部高校博士学科点专项基金项目(20070290516);教育部留学归国人员科研启动基金项目
作者(Author): 王小美,逄云峰,杜培军,谭琨,李光丽
DOI: 10.16251/j.cnki.1009-2307.2011.02.024
参考文献(References)：
- [1]Vapnik V N.The Nature of Statistical Learning Theory[M].New York:Springer-Verlag,1995.
- [2]Mercier,G,Lennon,M.Support Vector Machines forHyperspectral Image Classification with Spectral-basedkernels[C]//IEEE International Geosciences and Re-mote Sensing,2003,1:288-290.
- [3]Mahesh Pal,Paul M Mather.Support Vector classifiersfor Land Cover Classification[EB/OL].Image Pro-cessing&Interpretation Ma PIndia Conference,2003.http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.120.199&rep=rep1&type=pdf
- [4]Hsu C W,Lin C J.A comparison of methods for multi-class support vector machines[J].IEEE Transactionson Neural Networks,2002,13(2):415-425.
- [5]Gualtieri J Anthony,Crom PR F.Support Vector Ma-chine for Hyperspectral Remote Sensing Classification[C]//Proceedings of SPIE-The International Society ofPhoto Optical,1998,3584:221-232.
- [6]胡自申,张迁.基于SVM的遥感影像的分类[J].遥感信息,2003,(1):14-18.
- [7]Simon Perkins,Neal R Harvey,Steven P,et al.Sup-port Vector Machine for Broad Area Feature Classifica-tion in Remotely Sensed Images[C]//Proceedings ofSPIE-The International Society for Optical Engineering,2001,286-295.
- [8]Zhang J,Zhang Y,Zhou T.Classification of hyper-spectral data using support vector machine[C]//IEEE International Conference on Image Processing,2001:882-885.
- [9]常威威.高光谱图像条带噪声消除方法研究[D].西安,西北工业大学,2007.
- [10]杨金红,顾松山,程明虎.插值法在去除MODIS遥感影像条带噪声中的应用[J].气象科学,2007,27(6).
- [11]蒋耿明,牛铮,阮伟利,等.MODIS影像条带噪声去除方法研究[J].遥感技术与应用,2003,18(6).
- [12]许建华,张学工,李衍达.支持向量机的新发展[J].控制与决策,2004,19(5).
- [13]唐宏,等.光谱角制图模型的误差源分析与改进算法[J].光谱学与光谱分析,2005,25(8).
- [14]Giovani Corsini,Marco Diani,Tomas Walzel.StripingRemoval in MOS-B Data[J].IEEE Transaction onGeoscience and Remote Sensing,2000,38(3).
- [15]Lian Tzu-Chao,Yu Pao-Ta.Salt-Pepper impulse noisedetection and removal using multiple thresholds for im-age restoration[J].Journal of Information science andengineering,2006,22:189-198.
- [16]Bar,L,Sochen,N,Kiryati N.Image Deblurring inthe presence of salt-and-pepper noise[J].Interna-tional Journal of Computer Vision,2006,7(3).
- [17]Chan R H,Chung-Wa Ho and Nikolova M.Salt-and-pepper noise removal by median-type noise detectorsand detail-preserving regularization[J].IEEE Trans-actions on Image Processing,2005,14(10).