SAR图像弱反射地物边缘提取方法

Method of Edge Extraction of Small Reflection Object in SAR Image

提出了一种抑制ＳＡＲ图像相干斑噪声和提取弱反射地物边缘方法．传统噪声抑制和小波变换去噪方法有其不足之处，本文将小波变换和维纳滤波结合抑制ＳＡＲ图像相干斑噪声，通过选择恰当小波基获得良好滤波效果。采用最小错误准则计算ＳＡＲ图像的理论分割阈值，通过逐次迭代得到其合理大小，利用形态算子作用于分割图像获得其边缘．实际ＳＡＲ图像测试结果表明了本文方法的有效性．

De-noising for SAR image speckle and segmentation and edge extraction of small reflection object in SAN image is proposed in this paper. Traditional methods and recently wavelet transform for reducing speckle noise have some shortcoming,this study has satisfied result for reducing speckle noise by combining wavelet trans form and Wiener filter. The theory threshold for SAR image Segmentation is deduced by the criteria of smallest er ror, which is obtained by the means of iterative. Edge appears by morphological operating on segmented SAR image. The validity of those methods ate proved by experiments for real SAR image