论文部分内容阅读
为了直接利用超宽带(UWB)时域测量数据,同时重建二维(2D)目标(OI)的介电常数和电导率,本文将频域高斯-牛顿反演(GNI)算法发展为时域形式.迭代重建过程中,正问题由时域有限差分(FDTD)法求解,而逆问题的病态特性用自适应正则化技术抑制.四类数值算例中,噪声影响均被考虑,仿真结果初步证实了改进算法的可行性和鲁棒性.重建图像呈现超分辨率(SR),有望应用到早期乳腺癌检测等实际问题中.
In order to directly use UWB time-domain measurement data and reconstruct the permittivity and conductivity of two-dimensional (2D) target (OI) at the same time, this paper develops the frequency domain Gauss-Newton inversion (GNI) In the process of iterative reconstruction, the positive problem is solved by finite difference time-domain (FDTD) method, while the ill-posed characteristic of the inverse problem is suppressed by adaptive regularization technique.In the four numerical examples, the influence of noise is considered and the simulation results are preliminary confirmed In order to improve the feasibility and robustness of the algorithm, the reconstructed image is super-resolution (SR) and is expected to be applied to the practical problems of early breast cancer detection.