Regularization Imaging Algorithm with Accurate G Matrix for Near-Field MMW Synthetic Aperture Imaging Radiometer

In order to improve the reconstruction accuracy of near-fleld SAIR, a novel regularization imaging algorithm based on an accurate G matrix is proposed in this paper. Due to the fact that the regularization reconstruction is usually an underdetermined problem, inaccurate operation matrix G will lead to great reconstruction error in the imaging results, or even the normal imaging cannot be obtained. In this paper, we establish an accurate G matrix based on the accurate imaging model of near-fleld SAIR. Compared with the traditional G matrix with some unnecessary approximations, the proposed G matrix without approximation can improve the reconstruction accuracy efiectively. For improving the accuracy of matrix G further, the corresponding parameters are corrected according to the RMSE between the imaging results of the regularization method and modifled FFT method which is not sensitive to the parameters' change. The efiectiveness of this calibration method has been tested by 1D simulation experiments. Moreover, the 2D simulation experiments demonstrate that the proposed accurate G matrix can improve the imaging accuracy of regularization method efiectively. Finally, the 1D imaging experiment is performed to test the efiectiveness of the proposed method for the actual synthetic aperture imaging further.