基于GMRF-SVM的高分辨率遥感影像目标区域划分方法Research on region partition in high resolution remote sensing image based on GMRF-SVM
明冬萍,骆剑承,沈占锋
摘要(Abstract)：
高分辨率遥感影像数据量大、细节丰富并呈现出一定的尺度依赖性,单一尺度遥感影像分割难以同时兼顾影像的宏观和微观特征,这成为制约遥感信息自动化提取技术发展的瓶颈之一。对此本文提出了基于特征的多尺度高分辨率遥感信息提取技术框架,并分析了其对于大尺度海量数据信息提取与目标识别工作具有的理论及实践意义。根据影像光谱或纹理等特征,提出采用GMRF-SVM方法在大尺度上进行分类的目标区域划分方法。从大尺度信息提取的角度来看,该方法综合了GMRF纹理分类和SVM少量样本模式识别的优势,便于先验知识的融合,无论从花费时间还是分类处理效果上,都远远优于直接采用GMRF进行分割所取得的效果,对于后面的信息提取和目标识别来说更具有实际意义。
关键词(KeyWords)： 高分辨率遥感;信息提取;目标识别;多尺度;区域划分;GMRF-SVM
基金项目(Foundation): 国家自然科学基金项目(40601057)
作者(Author): 明冬萍,骆剑承,沈占锋
参考文献(References)：
- [1]明冬萍,等.高分辨率遥感影像信息提取与目标识别技术研究[J].测绘科学,2005,30(3):18-20.
- [2]沈庭芝,方子文.数字图像处理及模式识别[M].北京:北京理工大学出版社,1998.107-115.
- [3]Hideki N,Mahdad N S,Eiji K.MRF-based Texture Segmentation using Wavelet Decomposed Images[J].Pattern Recognition,2002,35(4):771-782.
- [4]Tseng D C,Lai C C.A Genetic Algorithm for MRF-based Segmentation of Multi-spectral Textured Images[J].Pat-tern Recognition Letters.1999,20(14):1499-1510.
- [5]Blake A,Zisserman A.Visual Reconstruction[M].Cambridge,MA:The MITPress,1987.
- [6]Chellappa R.Two-Dimensional Discrete Gaussian Markov Random Field Models for Image Processing[C]//Kanal L N,Rosenfeld A.Progress in Pattern Recognition.Amster-dam:Elsevier Science Publishers BV,1985,2:79-112.
- [7]Manjunath B S,Chellappa R.Unsupervised Texture Seg-mentation Using Markov Random Field Models[J].IEEE Transactions on Pattern Analysis and Machine In-telligence,1991,13(5):478-482.
- [8]Burges C J C.A Tutorial on Support Vector Machines for Pattern Recognition[J].Data Mining and KnowledgeDiscovery,1998,2:121-167.
- [9]Scholkopf B,Burges C J C,Smola A J.Advances in Kernel Methods:Support Vector Learning[M].Cam-bridge,MA:The MITPress,1999.
- [10]Cecilio A,Andreu C K.A Multi-class Support Vector Machine[C]//ECML2000,LNAI,2000,1810:31-38.
- [11]Cesmeli E,Wang D L.Texture Segmentation Using Gaussian Markov Random Fields and Neural Oscillator Networks[R].Techincal Report:OSU-CISRC-8/97-TR40,1997.
- [12]Huang Y,Chan K L,Zhang Z H.Texture Classification by Multi-model Feature Integration using Bayesian Net-works[J].Pattern Recognition Letters,2003,24:393-401.
- [13]Gold C,Sollich P.Model selection for support vector machine classification[J].Neurocomputing,2003,55:221-249.
- [14]Xin D,Wu Z,Pan Y.Probability Output of Multi-class Support Vector Machines[J].Journal of Zhejiang Uni-versity Science,2002,3(2):131-13.
- [15]明冬萍,骆剑承,周成虎,等.基于简化随机场模型的高分辨率遥感影像分割方法[J].计算机工程与应用,2004,40(26):28-30.