Journal of Systems Engineering and Electronics ›› 2011, Vol. 33 ›› Issue (9): 1983-1987.doi: 10.3969/j.issn.1001-506X.2011.09.13

• 传感器与信号处理 • 上一篇    下一篇

PolSAR图像的两级分块最优加权滤波方法

韩萍1, 晏珂2, 吴仁彪1,3
  

  1. 1. 中国民航大学智能信号与图像处理天津市重点实验室, 天津 300300
    2. 中国民航江西空管分局, 江西 南昌 330114
    3. 深圳大学ATR国防科技重点实验室智能信息处理研究室, 广东 深圳 518060
  • 出版日期:2011-09-17 发布日期:2010-01-03

Optimal weighting filtering based on twostage blocking for PolSAR images

HAN Ping1, YAN Ke2, WU Ren-biao1,3   

  1. 1. Tianjin Key Lab for Advanced Signal Processing, Civil Aviation University of China, Tianjin 300300, China
    2. Jiangxi Sub-bureau of Air Traffic Magement Bureau, Civil Aviation Administration of China, Nanchang 330114, China
    3. Intelligent Information Institute of ATR Lab, Shenzhen University, Shenzhen 518060, China
  • Online:2011-09-17 Published:2010-01-03

摘要:

针对多极化合成孔径雷达(polarimetric synthetic aperture radar, PolSAR)图像的相干斑抑制问题,提出一种两级分块的最优加权滤波方法。该方法在分块加权法的块处理基础上,采用边缘检测模板对非同质类子块进行二次分类和处理,提高了滤波参数的估计精度。实验结果表明,该方法不仅保持了块处理简单、运算量小的优点,而且在斑点抑制和边缘保持方面更优于最优加权法和分块加权法。

Abstract:

A method of optimal weighting filtering based on twostage blocking is proposed to solve the problem of speckle appearing in polarimetric synthetic aperture radar (PolSAR) images. On the basis of part weighting filtering, the method uses edge detecting template in each sub-image block to perform heterogeneous areas classification, so that more accurate filtering parameter estimation is achieved by calculating them with the classified area element values. Experimental results demonstrate that the proposed method not only keeps the advantages of blocking, but also achieves better performance both on speckle suppression and edge preserving compared with the optimal weighting and the part weighting methods.