基于滤波的SAR影像去噪方法分析

The Analysis of Noise Reduction for SAR Image Based on Filters

根据SAR(Synthetic Aperture Radar)影像相干噪声的特点,对现在广泛应用的各种抑制噪声的滤波方法如均值滤波、Frost滤波、增强型Lee滤波、Lee滤波、中值滤波等进行分析,并提出适合本研究区影像的基于小波变换的分量滤波处理去噪方法,利用均值、标准差、平滑指数、PM值、信息熵和平均梯度等评价指标对各种滤波去噪后的影像质量进行评价,并对去噪效果进行对比分析,证明本文所采用的滤波方法对SAR影像噪声有很好的抑制作用,并且较好的保存了图像纹理信息。

According to the characteristics of coherent noise in SAR （Synthetic Aperture Radar） images, we analysis variety of noise suppression filters those are widely used now, such as mean filter, Frost filter, Lee filter, enhanced Lee filter, Gamma Map filter and proposed subpart filter based on wavelet denosing method which is best in using in the study areas. Then, the mean values, standard deviation, smoothness index, PM values, standard deviation, the average gradient of entropy are used in evaluation of denoised images quality. The comparative analysis results show that the wavelet method used in this paper has a better inhibitory effect and preservation of image texture information.