基于聚类算法的SAR图像去噪  

SAR image denoising based on clustering algorithm

作　　者：路延[1] 

机构地区：[1]陕西职业技术学院,陕西西安710100

出　　处：《现代电子技术》2014年第8期126-128,共3页Modern Electronics Technique

摘　　要：遥感图像的用途非常广泛,而合成孔径雷达图像是遥感图像中重要的一种,人们能从中提供更多的有用信息,但其固有的相干斑给人们对其信息的提取带来了困难,去除相干斑成为SAR图像信息提取中最重要的一步。在此介绍了以模糊C均值聚类算法为基础,同时结合小波变换,对SAR图像进行去噪。并将实验结果与已有的SAR图像去噪方法进行实验、比较和分析。结果表明,模糊C均值聚类和小波变换的所结合的方法,在SAR图像去除斑噪的领域中,成为一种有效且吸引人的算法。The remote sensing images are widely used. Synthetic aperture radar （SAR） image is an important type of re-mote sensing images. However,the inherent speckle makes it difficult to extract the information form the image,so speckle de-noising becomes the most important step of information extraction. The SAR image denoising method based on fuzzy C-means clustering algorithm and combined with wavelet transform is introduced in this paper. Its experiment results are compared with those of several available methods. The result shows that the method combining fuzzy C-means clustering with wavelet transform is effective and attractive in the speckle denoising domain of SAR images.

关 键 词：合成孔径雷达 模糊C均值聚类 小波变换 图像去噪 

分 类 号：TN964-34[电子电信—信号与信息处理]