基于粒子群算法的全极化SAR图像非监督分类算法研究

基金项目: 国家863计划资助项目(2009AA12Z145)；国家自然科学基金资助项目(40523005)；武汉大学测绘遥感信息工程国家重点实验室专项科研经费资助项目

详细信息

Unsupervised Classification of Fully Polarimetric SAR Data Based on the PSO Algorithm

¹School of Remote Sensing and Information Engineering,Wuhan University,129 Luoyu Road,Wuhan 430079,China;(2 Chinese Academy of Surveying and Mapping,16 Beitaiping Road,Haidian District,Beijing 100039,China

Funds: 国家863计划资助项目(2009AA12Z145)；国家自然科学基金资助项目(40523005)；武汉大学测绘遥感信息工程国家重点实验室专项科研经费资助项目

摘要: 提出了一种基于H/α/A和粒子群优化(PSO)算法的全极化SAR数据非监督分类方法。该方法利用H/α/A对全极化SAR数据进行基于散射机理的初分类,计算各类别的聚类中心,并利用计算结果对PSO算法进行初始化,然后采用PSO对极化SAR数据进行迭代分类。在运算过程中,引入了基于最大似然准则的复Wishart距离,以提高分类器的性能。实验结果验证了该算法的有效性,所提出算法的分类结果优于传统的Wishart-H/α/A分类方法。
- 粒子群算法 /
- H/α/A /
- 极化SAR /
- 复Wishart距离
Abstract: A new unsupervised classification method of fully polarimetric SAR data based on the PSO algorithm and H/α/A is presented.Firstly,the result obtained by H/α/A classification is used to initialize the clustering centers,and then the fully polarimetric SAR data is classified by the PSO algorithm.Meantime,the Wishart distance measure is employed to improve the performance of the PSO-H/α/A classifier.Experimental results show that the new scheme proposed in this paper can effectively classify the fully polarimetric SAR data.
- PSO algorithm /
- H/α/A /
- polarimetric SAR /
- Wishart distance measure