基于边缘保持的自适应SAR图像超分辨重建算法

Adaptive Edge-preserving Based Reconstruction Algorithm for SAR Images

由于超分辨率图像的重建是典型的不适定问题,因此常引入正则化约束来稳定求解。在正则化方法的基础上,提出了一种针对SAR图像不同区域自适应处理的正则SAR图像超分辨重建新算法。该算法根据图像的邻域一致性测度构造自适应平滑约束项,并定义自适应的p L范数对此项进行度量。同时,构造基于1L范数的梯度逼近范数项,将其加入正则化的重建过程中,进一步提高SAR图像的重建质量。实验结果表明,该算法不仅具有良好的稳健性,而且在较好地重建SAR图像细节信息的同时,有效地抑制了斑点噪声,明显改善了重建SAR图像的质量。

The super resolution images reconstruction approach is an ill-posed problem, so regularization is often adopted to stabilize the inversion of this problem. Based on the framework of regularized super resolution （RSR） reconstruction approach, an adaptive edge-preserving RSR reconstruction algorithm for SAR images was proposed, which processed different regions of SAR images adaptively. In the adaptive super resolution reconstruction algorithm, an adaptive smoothness constraint term with adaptive Lp norm was introduced based on the neighborhood homogeneous measurement ~HM）. Meanwhile, the image gradient error term with L~ norm was combined in the regularized super resolution reconstruction to further improve the quality of the reconstruction SAR image. The experimental results validate the effectiveness of the proposed algorithm and demonstrate that it has the advantage of preserving detailed components and suppressing speckled noise.