针对SAR影像边缘检测受斑点噪声影响严重和极化信息利用不充分的问题,用滑动模板边缘两侧目标的协方差矩阵代替了极化白化滤波中杂波背景与窗口中心的协方差矩阵,提出一种基于改进极化白化滤波的边缘检测新方法,充分利用了极化通道间的相关性,在有效抑制斑点噪声的同时,提高了极化信息的利用率。模拟和真实极化影像的实验验证了新方法的有效性。

A new edge detector based on an improved polarimetric whitening filter (PWF) for polarimetric synthetic aperture radar (SAR) images is proposed,aiming at a better usage of polarimetric information and reducing the speckle noise. The covariance matrices of the background and the center in PWF are replaced by those at both sides of the moving edge template. Polarimetric information including complex correlation between different polarizations is better utilized while speckle is suppressed at the same time. The new method is demonstrated using simulated and real polarimetric SAR data.