摘要
本文介绍利用小波分析和软门限对合成孔径雷达（ＳＡＲ）图像斑点噪声（Ｓｐｅｃｋｌｅ）进行抑制与滤除的算法。首先选择合适的小波基对合成孔径雷达（ＳＡＲ）图像进行小波分解，应用ＤａｖｉｄＬ．Ｄｏｎｏｈｏ的软门限（ｓｏｆｔ－ｔｈｒｅｓｈｏｌｄｉｎｇ）理论，并且将其推广到２维信号。针对ＳＡＲ图像斑点噪声的特殊性，即在小波变换域内图像和斑点噪声的奇异性不同，选取合适的门限在小波域内滤波。最后进行小波反变换得到去噪后的ＳＡＲ图像。实验证明，该算法能有效的对合成孔径雷达ＳＡＲ图像斑点噪声滤除。
This paper presents a speckle reduction algorithm by Soft thresholding for SAR images in multiscale and multiorientation by using nolinear wavelet analysis. Firstly, a base is chosen according to SAR image features, which is used to make SAR images 2D wavelet transformation. Furthermore, we apply the David L. Donoho's soft thresholdning theory and extend to 2D SAR image processing, then select suitable soft threshold is aimed at SAR image features. After, remove speckles by 2D wavelet filtering. Lastly, invert the pyramid filtering recovering SAR image. The content experimental result of remove speckles on SAR images is expected.
出处
《测绘学报》
EI
CSCD
北大核心
1998年第2期119-124,共6页
Acta Geodaetica et Cartographica Sinica
基金
中国科学院自动化所模式识别国家重点实验室基金
国防科技预研基金
关键词
SAR图像
小波变换
斑点噪声滤除
软门限
SAR image, Wavelet transform, Speckle reduction, Soft thresholding
