基于区域划分的合成孔径雷达图像相干斑各向异性扩散滤波算法  

Anisotropic Diffusion Despeckling Algorithm for Synthetic Aperture Radar Images Based on Region Subdivision

作　　者：李晓东[1] 朱磊[1] 潘杨[1] 巩学芳 LI Xiao-dong;ZHU Lei;PAN Yang;GONG Xue-fang(College of Electronics and Information,Xi'an Polytechnic University,Xi'an 710048,China)

机构地区：[1]西安工程大学电子信息学院,西安710048

出　　处：《科学技术与工程》2018年第10期65-70,共6页Science Technology and Engineering

基　　金：国家自然科学基金(61401347);陕西省科技厅工业科技攻关项目(2016GY-101);陕西省教育厅自然科学基金(17JK0343)资助

摘　　要：经典的细节保持各向异性扩散滤波算法(DPAD)在抑制合成孔径雷达(SAR)图像乘性相干斑噪声时,存在同质区块效应现象与边缘区噪声残留问题;为此提出了一种基于区域划分的各向异性扩散滤波算法。该算法通过在DPAD扩散方程中耦合方向扩散来抑制边缘区域相干斑与同质区块状起伏。通过阈值化的边缘强度映射对SAR图像实施同质区与边缘区的区域划分;并对不同区域采用不同尺度局域窗分别实施各向异性扩散滤波,从而在抑制同质区相干斑的同时有效保护边缘。实验结果表明:与多种各向异性扩散抑斑算法相比,算法在保护边缘的同时,能明显降低同质区块效应现象与边缘区噪声残留问题。The classic detail-preserving anisotropic diffusion(DPAD)used for smoothing multiplicative speckle noise in SAR images suffers from blocking artefacts in homogeneous regions and poor despeckling performance near edges.In view of the above question,a new anisotropic diffusion filtering algorithm based on region division is proposed.The proposed algorithm is established to reduce speckle near edges and blocking artefacts in homogeneous regions by coupling directional diffusion equation into the DPAD s diffusion equation.The SAR image is divided into homogeneous regions and edge regions by the shreshold-processing edge strength maps.The anisotropic diffusion with different scale local windows in different regions is implemented,resulting in good speckle suppression in homogeneous regions and well detail protection in edge regions.The experimental results and comparisons with several traditional anisotropic diffusion algorithms show that the proposed algorithm can significantly reduce blocking artefacts in homogeneous regions and the residual noise in edge regions while protecting SAR images edges.

关 键 词：SAR图像 相干斑抑制 各向异性扩散 方向扩散 区域划分 

分 类 号：TN911.73[电子电信—通信与信息系统]