摘要
多视极化白化滤波器(MPWF)是一种专门应用于极化SAR图像降噪的有效方法,其中,滤波器参数估计的精确度直接决定了其滤波性能的好坏。本文通过对全极化SAR噪声模型的介绍和几种典型的参数估计方法的比较,针对传统方法的缺陷,提出基于无监督分类的自适应参数估计方法。此方法以分类图像作为对象;在滑动矩形窗内以中心像素作为参照物,自动搜索与其同类的像素并用于降噪。实验结果表明,该法不仅有效地抑制了相干斑,而且对图像的纹理信息具有很好的保持能力。
The multilook polarimetric whitening filter is an effective method on speckle reduction in multilook polarimetric synthetic aperture radar （SAR） images, where the function of the filter is directly decided by the precision of parameter estimation. In this paper, the model of the polarimetric synthetic aperture radar （SAR） is stated here, with some typical covariance matrix parameter estimate methods, which are applied to the multilook polarimetric whitening filter. Aimed at traditional defects, a novel approach based on unsupervised classification is proposed here, where the classified image is chosen as the processed object and the central pixel in moving rectangular window is chosen as reference; and then through automatic search, the pixels in the same class are selected and used for despeckling. The experimental results demonstrate the effectiveness both on speckle reduction and preservation of texture information from the experimental- results. Comparisons are also made.
出处
《电子测量技术》
2006年第6期71-72,83,共3页
Electronic Measurement Technology
关键词
极化合成孔径雷达
相干斑
无监督分类
MPWF
polarimetric synthetic aperture radar
speckle
unsupervised classification
MPWF
