Research On Spatial Resolution Enhancement Method Of Spaceborne Microwave Radiometer Based On Convolution Inversion

Microwave remote sensing imaging technology plays an important role in the field of atmospheric,oceanic and terrestrial observation and global environmental monitoring and protection,and has urgent needs in the field of military detection and other applications.Passive microwave remote sensing radiometer are acquired by using mechanical scanning of antenna for observation field of view scene information,which is not dependent on light conditions,not transmitting,the advantages of the strong penetrating power,but the traditional real aperture antenna radiometer need large size in remote sensing satellite load significantly,and compared with visible light,the imaging resolution is too low.How to establish the imaging detection system and improve the brightness temperature image resolution by inversion technology is a hot issue that needs to be studied.In this thesis,based on SMOS satellite working mode,a comprehensive aperture radiometer detection model is established,and constrained least squares filter and derived constrained least squares filter are proposed to accurately estimate the brightness temperature grid of target features,thus achieving better resolution of brightness temperature data products.In addition,a three-stage truncated singular value decomposition（TSVD）method is proposed to optimize the resolution of irregular samples.The main contents are as follows:1.The geometric configuration of synthetic aperture microwave radiometer is established and the brightness temperature inversion model of synthetic aperture radiometer is studied.The brightness and temperature images of SMOS radiometer data were extracted by windowing-adding and redundant equalization pretreatment techniques,which laid a theoretical foundation for subsequent high-resolution processing.2.A deconvolution filtering method in frequency domain was proposed,and the best approximation of the target signal function was established based on the Lagrange multiplier method.The deconvolution restoration effect with higher spatial resolution was obtained by adjusting the linear operator,and the brightness temperature resolution was improved.3.In order to solve the problem of irregular interval of radiometer sampling data,a regularized irregular discrete sampling method is proposed,and the problem of large inversion error caused by irregular interval of radiometer sampling data is solved by using singular value inverse filtering method to convolution matrix.The microwave radiometer imaging inversion model and the method to improve brightness temperature image resolution mentioned above have been verified by computer simulation and SMOS satellite radiometer data,and the effectiveness of the method has been proved.