Research On Image Reconstruction Method Of Synthetic Aperture Microwave Radiometer

As a new technology in the field of passive microwave remote sensing,the synthetic aperture microwave radiometers use synthetic aperture interferometric measurement technology,which are easy to achieve higher spatial resolution.Simultaneously,they also have the advantages of rapid imaging in a large field of view,no mechanical scanning,and have important application value in land,ocean and atmosphere remote sensing and other fields.Inversion imaging is a key content of synthetic aperture microwave radiometer.The inversion imaging process of synthetic aperture interferometric radiometers(SAIRs)is to transform the measured visibility function data into radiant brightness temperature distribution,which is a ill-conditioned inverse problem,and its solution is neither unique nor stable.Although the traditional regularization methods can effectively overcome the ill-posed nature of the inverse problem,there are still problems such as large image reconstruction errors or slow computing speed,and the accuracy or speed of the inverse imaging can not meet the requirements.Therefore,this paper proposes several new synthetic aperture interferometric radiometers image reconstruction algorithms to improve the accuracy or speed of inversion imaging.The main research content is as follows:(1)Although the traditional iterative regularization method has good smoothing effect and robustness,it has the problem of slow computation speed.Therefore,this paper proposes a synthetic aperture interferometric radiometer imaging method based on accelerated iterative regularization.This method utilizes a negative penalty term to modify the traditional least squares function to accelerate the initial iteration process.Concurrently,a series of decreasing coefficients need to be set to prevent noise amplification during subsequent iterations.The simulation results show that compared with the traditional iterative regularization,The accelerated iterative regularization method can significantly reduce the number of iterations without reducing the reconstruction accuracy,thus effectively improving the imaging speed of synthetic aperture interferometric radiometer.(2)In order to reduce the reconstruction error of traditional regularization,a synthetic aperture interferometric radiometer imaging method based on reweighted total variation regularization is proposed.Simultaneously,the Split Bregman iterative algorithm is used to optimize the solution of the reweighted total variation regularization model to improve the calculation speed.The simulation results show that,compared with the traditional minimum-norm regularization and band-limited regularization,this method can effectively reduce the Gibbs oscillation error,retain the edge information of the image,and thus effectively improving the accuracy of synthetic aperture interferometric radiometer inversion imaging.(3)A synthetic aperture interferometric radiometer imaging method based on adaptive L~pis proposed to reduce the over smooth effect and fluctuation oscillation existing in the classical Hilbert space regularization method.This method combines the advantages of Hilbert space and Banach space regularization,and can adaptively adjust the p value of L~pnorm between 1.2 and 2 according to the characteristics of the brightness temperature map to be reconstructed.The simulation results show that,compared with the minimum-norm regularization and band-limited regularization,this method can effectively reduce the smoothness and Gibbs oscillation of reconstruction results,and thus effectively improving the accuracy of synthetic aperture interferometric radiometer inversion imaging.