摘要
提出一种基于聚类集成的半监督多/高光谱图像分类方法。谱聚类是近年来出现的基于图论的、以相似性为基础的一类性能优越的聚类算法,能对任意形状分布的数据进行聚类,但对参数的变化比较敏感。聚类集成技术可有效提高单聚类算法的精度和稳定性,并具有良好的鲁棒性和泛化能力。算法利用聚类集成算法的优点并利用谱聚类的思想开发聚类集成算法的共识函数,将谱聚类作为聚类成员来构造聚类集成系统,使用高斯RBF核映射下的多维数据的光谱角制图计算权值矩阵W,并用Nystrm方法来降低算法的运算复杂度,实现了多/高光谱遥感数据的半监督分类。最后通过实验验证了该算法无论对多光谱还是高光谱都有较好的分类效果。
A semi-supervised, cluster-ensemble based method for the classification of multi / hyperspectral images is presented. Spectral clustering is a graph theory based clustering algorithm taking similarity as the basis, and has become increasingly popular in recent years. It can deal with arbitrary distribution of dataset but with a drawback for being sensitive to the scaling parameters. Cluster ensemble techniques are effective in improving both the robustness and the stability of the single clustering algorithm. Cluster ensemble also has a character of good robustness and generalization ability. The processing method in this paper utilizes the merits of cluster ensemble and develops a consensus function based spectral clustering algorthm. The clustering components are generated by spectral clustering. The affinity matrix is generated by computing the SAM between different datapoints. The Nystrm method is used to to speed up the classification process.Thus semi-supervised classification to multi / hyperspectral remote sensed data is completed. Experiments show that the method presented here has an excellent classificaation result for both multispectral and hyperspectral remote sensed dataset.
引文
[1]MATHER P M,KOCH M.Computer processing of remotely-sensed images[M].4th ed.New York:A John Wiley&Sons,Ltd.,Publication,2011:229-284.
[2]ALAJLAN N,BAZI Y,MELGANI F,et al.Fusion of supervised and unsupervised learning for improved classification of hyperspectral images[J].Information Sciences,2012,217:39-55.
[3]罗小波,赵春晖,潘建平,等.遥感图像智能分类及其应用[M].北京:电子工业出版社,2011:55-57.(LUO X B,ZHAO C H,PAN J P.et al.Remote sensing image intelligent classification and its application[M].Beijing:Publishing House of Electronics Industry,2011:55-57.)
[4]JOHNSON B,XIE Z X.Unsupervised image segmentation evaluation and refinement using a multi-scale approach[J].ISPRS Journal of Photogrammetry and Remote Sensing,2011,66(4):473-483.
[5]DUNDAR M M,LANDGREBE D A.A cost-effective semisupervised classifier approach with kernels[J].IEEE Transactions on Geoscience and Remote Sensing,2004,42(1):264-270.
[6]王娜,李霞.基于监督信息特性的主动半监督谱聚类算法[J].电子学报,2010,38(1):172-176.(WANG N,LI X.Active semi-supervised spectral clustering based on pairwise constraints[J].Acta Electronica Sinica,2010,38(1):172-176.)
[7]王雪松,张晓丽,程玉虎,等.一种基于谱聚类的聚类核半监督支持向量机[J].中国矿业大学学报,2010,39(6):886-890.(WANG X S,ZHANG X L,CHENG Y H,et al.A cluster kernel semi-supervised support vector machine based on spectral clustering[J].Journal of China University of Mining&Technology,2010,39(6):886-890.)
[8]MANTRACHA A,VAN ZEEBROECK N,FRANCQ P,et al.Semi-supervised classification and betweenness computation on large,sparse,directed graphs[J].Pattern Recognition,2011,44:1212-1224.
[9]CAMPS-VALLS G,BANDOS T,ZHOU D Y.Semi-supervised graph-based hyperspectral image classification[J].IEEE Transactions on Geoscience and Remote Sensing,2007,45(10):3044-3054.
[10]刘汉强,赵凤.基于空间特征的谱聚类含噪图像分割[J].模式识别与人工智能,2012,25(3):419-425.(LIU H Q,ZHAO F.Space feature based spectral clustering for noisy image segmentation[J].PR&AI,2012,25(3):419-425.)
[11]马秀丽,焦李成.基于分水岭-谱聚类的SAR图像分割[J].红外与毫米波学报,2008,27(6):452-456.(MA X L,JIAO L C.SAR image segmentati on based on watershed and spectral clustering[J].J.Infrared Millim.Waves,2008,27(6):452-456.)
[12]REBAGLIATI N,VERRI A.Spectral clustering with more than K eigenvectors[J].Neurocomputing,2011,74(9):1391-1401.
[13]周林,平西建,徐森,等.基于谱聚类的聚类集成算法[J].自动化学报,2012,38(8):1335-1342.(ZHOU L,PING X J,XU S,et al.Cluster ensemble based on spectral clustering[J].Acta Automatica Sinica,2012,38(8):1335-1342.)
[14]STREHL A,GHOSH J.Cluster ensembles-a knowledge reuse framework for combining partitions[J].Journal of Machine Learning Research,2002,3:583-617.
[15]TUMER K,AGOGINO A K.Ensemble clustering with voting active clusters[J].Pattern Recognition Letters,2008,29(14):1947-1953.
[16]MIMAROGLU S,ERDIL E.Combining multiple clusterings using similarity graph[J].Pattern Recognition,2011,44:694-703.
[17]YU Z W,WONG H S,YOU J,et al.Hybrid cluster ensemble framework based on the random combination of data transformation operators[J].Pattern Recognition,2012,45:1826-1837.
[18]VANDER MEER F.The effectiveness of spectral similarity measures for the analysis of hyperspectral imagery[J].International Journal of Applied Earth Observation and Geoinformation,2006,8(1):3-17.
[19]LANDGREBE D.Hyperspectral image data analysis[J].Signal Processing Magazine,IEEE,2002,19(1):17-28.
[20]MONTSERUD R A,LEAMANS R.Comparing global vegetation maps with the kappa statistic[J].Ecological Modeling,1992,62(4):275-293.
[21]高恒振.高光谱遥感图像分类技术研究[D].长沙:国防科学技术大学,2011.(GAO H Z.Research on classification technique for hyperspectral remote sensing imagery[D].Changsha:National University of Defense Technology,2011.)