The Research Of Speckle Reduction Of Polarimetric Synthetic Aperture Radar

A polarimetric SAR combines both advantage of measuring full complex scattering matrix for each resolution cell in a scene and producing two-dimensional high resolution images, and is thus playing a more and more important role in gathering the detailed information regarding the geometric and electric structures of surfaces and covers for application in both civil and military areas.Because speckle in the SAR or polarimetric SAR images badly affects the post-processing of SAR images, speckle reduction become to the first step of SAR images post-processing. The research project of the dissertation is the national natural science fund project. The main target of this project is to suppress the speckle of SAR or polarimetric SAR images. Regarding to the algorithms of speckle reduction, many algorithms have been presented abroad and domestic. For examples, the multi-looks averaging, the filtering of spatial domain and the filtering of polarimetric domain and so on.The main content of this dissertation are as follow:1)The author analyses and derives the statistics characteristic of speckles, including the statistics characteristic of speckles of the intensity or amplitude images in the spatial domain and the one of speckle vector and speckle matrix in the polarimetric domain.2)The author analyses the multiplicative speckle noise model in the spatial domain and the polarimetric domain, and imrove the polarimetric multiplicative speckle noise model to meet the practicality.3)The author analyses and derives the more famous despeckle algorithms, for examples, J.S. Lee抯 local statistics filter, MAP filter in the spatial domain, L. M. Novak and M. C. Burl抯 polarimetric whitening filter, Lee抯 optimal weighting algorithm in the polarimetric domain and so on. The author has already made a program to implement these algorithms.4)On the base of improved polarimetric multiplicative speckle noise model, the author presents a new despeckle algorithm--multi-texture maximum likelihood estimator, which estimates simultaneously the texture parameters of multi-polarimetric channels by the method of maximum likelihood estimation and so get the despeckled SAR intensity images of multipolarimetric channels. The experiments indicated that this algorithm got the better effect of speckle reduction.Furthermore, the author has analyses the SIR-C CEOS polarimetric SAR image data format which is very complicate and made a program to decompress the image data for the successor to use.The application of polarimetric SAR remote sensing in our country has newly begun. The investigation in the field of polarimetric SAR undoubtedly has a bright future in application. The research performed and achievement obtained in the dissertation, which consummate the theory of polarimetric SAR signal processing and enhance the performance of polarimetric SAR application, are of signification in both theory and practice, and will be certainly making an active contribution to expediting and improving the applications of polarimetric SAR remote sensing for our country.