摘要
极化合成是极化SAR图像处理的一种重要方法,它能在成像处理后,利用已获得的Sinclair矩阵重新生成任意极化方式下的雷达接收功率图像,并能通过选取收发天线极化状态相同或正交,分别得到描述目标散射特性的共极化特征图和交叉极化特征图。根据极化合成理论和极化特征图的概念,可以获取目标的最佳极化。将其作为分类器的输入特征量,提出了一种基于极化合成的目标分类算法,并对实测极化SAR数据进行了分类实验。结果表明,该算法对于从极化SAR数据中获取目标的最佳极化,进而对目标进行分类是可行和有效的。
Polarization synthesis is an important way to polarimetric SAP, image processing. It can get receiver power image in arbitrary polarization states using Sinclair matrix after imaging processing. By selecting different antenna states,co- polarization and cross- polarization signature can represent target scattering characteristic. Optimum polarization of targets can be obtained based on polarization synthesis theory and polarization signatures. Algorithm of target classification based on polarization synthesis is proposed, when it is acted as input value of classifier. Then,polarimetric SAR data is applied to classification experiment. The results indicate this algorithm is feasible and efficient to obtain optimum polarization of targets from polarimetric SAR data and classify targets.
出处
《计算机技术与发展》
2007年第3期30-32,36,共4页
Computer Technology and Development
基金
国家自然科学基金资助项目(69971001)
关键词
极化合成
最佳极化
目标分类
最小距离分类器
polarization synthesis
optimum polarization
target classification
minimum distance classitier
