在过去的 2 0多年中 ,提出了许多去除SAR图像斑点噪声的方法。这些方法的主要目的是去除斑点噪声而不破坏图像的空间分辨率和边缘信息。理论和实践表明去除图像的斑点噪声和保持图像边缘信息是无法同时完成的。现有大部分滤波器是在去除噪声和保持边缘之间进行折衷。在Han等人方法的基础上提出了一种在去除噪声和保持边缘方面最优的滤波器。它在去除噪声同时能够有效保持边缘信息

Speckle in synthetic aperture radar (SAR) images disturbs SAR images calibration, validation, and interpretation. Thus in most cases of practical interest, reducing speckle in SAR images is a goal towards which we aspire. Over the past 20 years, numerous methods to reduce speckle have been proposed based on digital image processing after the images have been formed. A number of filters have been proposed to reduce speckle based on speckle models, such as the Lee, the Kuan, the Frost, the modified Lee, modified Frost, and filters based on the Maximum A Posteriori (MAP) probability. Meanwhile other filters are not based on speckle models, such as median filter, geometric filter, and wavelet transform filter but also have been applied to reduce speckles. The primary goal of these methods is to reduce speckle without destroying resolution and edge information. In fact, the experiments indicate that there is always a kind of tradeoff between smoothing out speckle and preserving the useful information. It is an inverse problem to restore images with speckle on mathematical physics view. The common attribute of inverse problem is ill-conditioned. Thus, it is impossible to obtain the true image from the images with speckle. Even so, approximate image can be restored from images with speckle on certain condition. In fact, many methods to restore image with speckle have been proposed based on different principle. The different images can be acquired from the same image with speckle using different filters. This makes a problem that which image is optimal. In the general, an optimal filter is well-balanced in the quality of visual appearance, mean preservation, edge preservation, and reduction of the standard deviation. In this paper, an enhanced SAR speckle filter is developed. It can effectively smooth out speckle while preserve edge information.