SAR Imagery Classification Using Multi-Class Support Vector Machines

摘要：

In this paper,we present a Multi-class Support Vector Machines (M-SVM) application to remote-sensing SAR image classification. M-SVMs are an n-ary extension of Support Vector Machines (SVM),introduced by Vapnik within the framework of the Statistical Learning Theory. In this article we use the M-SVMs in order to classify a ERS-1 SAR multi-frequency survey of Torre de Hercules coast,Spain (December 13,1992),preprocessed by a gray-level scaling thanks to a selfimplemented Matlab function. Main objective of this work is evaluate the classification performances of M-SVMs in comparison with most frequently employed Neural Network and Fuzzy classifiers. The proposed algorithm returned interesting results with respect to Neural Network and Fuzzy classifiers,having a reliability factor around to 94％.

作者: G.Angiulli V.Barrile M.Cacciola

作者单位: University “Mediterranea of Reggio Calabria,Italy

会议类型: 国际会议

会议名称: Progress in Electromagnetics Research Symposium 2005(2005年电磁学研究新进展学术研讨会)(PIERS 2005)

会议地点: 杭州

会议语种:英文

页码: 218-222

在线出版日期: 2005-08-22（万方平台首次上网日期，不代表论文的发表时间）