基于小波变换的高分辨率SAR图像相干斑噪声抑制研究

Research on Coherent Speckle Reduction in High Resolution SAR Image Based on Wavelet Transform

论述了基于小波变换对高分辨率合成孔径雷达(SAR)图像的相干斑噪声进行抑制。首先从SAR图像相干斑噪声的产生机理出发,论述了在SAR图像中采用小波变换对相干斑噪声滤波的优越性和必要性;然后详细论述了在小波变换域中如何对相干斑噪声进行有效的滤波,同时利用小波细节子图像提供的信息来检测边缘和纹理细节;最后通过实验结果说明了此方法不但对高分辨率SAR图像相干斑噪声有良好的抑制作用,而且还尽可能多地保留了目标特性和纹理细节,使图像的视觉特性良好。

This paper deals with coherent speckle reduction in high resolution SAR image based on wavelet transform. First of all, the reason of generating speckle in SAR image is interpreted, the advantage and the necessity of speckle reduction based on wavelet transform are discussed. And then, it is expounded in detail how to filter for reducing speckle in wavelet transform domain, while the information contained in subband images is utilized for detection edge and texture detail. Finally, the application of the method to high resolution SAR images have shown that the performance of the filter is satisfactory in speckle reduction and edge preservation, and in generating visual images as well.