SAR图像相干斑噪声滤除中局部窗的选择

    Method of Selecting the Local-windows in SAR Image Speckle Filtering

    • 摘要: 许多合成孔径雷达(SAR)图像相干斑噪声滤除方法中都需要用到SAR图像的局部统计特性,这就不可避免地要用到滑动局部窗。通常都是用边长为奇数像素的正方形窗作为局部窗,但这些窗的使用均造成了滤波之后图像边缘及细节的模糊甚至丢失。提出了一种基于保留边缘及细节的包含子窗的非正方形局部窗,能够在滤除相干斑的同时很好地保持图像的边缘及细节纹理信息,并且计算量小,易实现,有较好的实用性。

       

      Abstract: The local statistical properties of synthetic aperture radar(SAR) image are used in many speckle filtering methods,and a boxcar filter must be used.Generally,a square window with odd number of pixels width is used as the local window,but it may smear edges and details.The nonsquare edge-aligned local windows with subwindows are proposed,which can despeckle and also preserve image edges and details.Besides,they have low computation complexity,easiness of realization and better practicability.

       

    /

    返回文章
    返回