论文标题
SAR图像使用Wirtinger微积分和Cauchy正则化自动关注
SAR Image Autofocusing using Wirtinger calculus and Cauchy regularization
论文作者
论文摘要
在本文中，提出了使用Cauchy正则化的优化模型，以同时进行SAR图像重建和自动焦点。坐标下降框架，其中对所需图像和相位误差进行了优化，以解决模型。对于估计图像的子问题，我们利用Wirtinger演算的技术直接最大程度地减少涉及复杂变量的成本函数。我们还利用了最先进的稀疏性辅助凯奇正规器。通过在模拟场景和真实的SAR图像上进行实验，提出的方法被证明可以给出令人印象深刻的自动关注结果。
In this paper, an optimization model using Cauchy regularization is proposed for simultaneous SAR image reconstruction and autofocusing. A coordinate descent framework in which the desired image and the phase errors are optimized alternatively is designed to solve the model. For the subproblem of estimating the image, we utilize the techniques of Wirtinger calculus to directly minimize the cost function which involves complex variables. We also utilise a state-of-the-art, sparsity-enforcing Cauchy regularizer. The proposed method is demonstrated to give impressive autofocusing results by conducting experiments on both simulated scene and real SAR image.
