基于小波系数统计特征的SAR图像恢复

基金项目: 国家自然科学基金资助项目(40376051)

详细信息

SAR Image Restoration Based on Statistic Characteristics of Wavelet Coefficients

¹ School of Electronic Information,Wuhan University,129 Luoyu Road, Wuhan 430079,China

Funds: 国家自然科学基金资助项目(40376051)

摘要: 基于小波系数统计特征,提出了一种SAR图像恢复方法。首先将对数变换后的SAR图像进行平稳小波分解,基于小波系数统计特征,用-αstable分布分别对纯净信号和噪声建模,估计出各部分模型参数后,用MAP估计器对处理后的SAR图像进行非线性处理来恢复出纯净图像。实验结果表明,该方法能在抑制噪声的同时最大程度地保持纹理细节。
- 合成孔径雷达 /
- -αstable分布 /
- 小波变换 /
- 相干斑
Abstract: An algorithm within framework of wavelet analysis based on the statistic characteristics of wavelet coefficients is proposed.The wavelet decomposition of logarithmically transformed SAR images is accurately modeled by stable distributions.The result of experiment indicates that the proposed method reduces speckle in SAR images while preserving the structural feature and textural information of the scene.
- synthetic aperture radar /
- α-stable distribution /
- wavelet decomposition /
- speckle