Study On Multibaseline Tomography Synthetic Aperture Radar

Multibaseline tomography synthetic aperture radar(SAR) is an extension ofinterferometric SAR(InSAR). Multibaseline tomography SAR synthesizes a very longaperture along the height direction via increasing many antennae on the basis ofInSAR. Multibaseline tomography SAR has the ability of three-dimensional imaging.As a new technique, multibaseline tomography SAR shows huge potentialities inapplications spanning geology, forestry, biomass estimation, archeology, glacier anddetection of buried structures. This dissertation studies the multibaseline tomographySAR intensively.In this dissertation, the primary aspects of single baseline InSAR is investigated,which involves the basic principle, the accuracy of elevation; the decorrelation anddata processing. The coregistration of complex image of InSAR is specially studieddetailedly. The three-dimensional imaging techniques of multibaseline tomographySAR is also studied intensively in this dissertation.This dissertation is devoted to:a) The averaged periodogram of phase interferogram is proposed as theevaluation function for coregistration of InSAR complex images. Theestimation variance of evaluation function is reduced by separating the datainto numbers of subsections and averaging them, which can improve theprecision and reliability of coregistration.b) The height-dimensional imaging method based on modem spectrumestimation techniques is proposed for multibaseline tomography SAR. Thenew method uses the power spectrum as the height-dimensional image oftomography SAR and the power spectrum of data measured is estimated bymodern spectrum estimation techniques. This imaging method can overcomethe problem that there are only few baselines in tomography SAR systemsand can get very high resolution in the height dimension. c) The height-dimensional imaging method based on anay signal processingtechniques is proposed for tomography SAR. This imaging method uses theangular spectrum as the height-dimensional image of tomography SAR. Itcan breakthrough the Rayleigh limitation and get great higher resolution thanthe Fourier transform method.d) The height-dimensional imaging scheme based on BG inverse algorithm isproposed. The nouniform baselines will bring serious bad influence to theimaging quality of tomography SAR, the BG inverse algorithm can transformthe non-orthogonal kernel functions into orthogonal ones and reconstructs theobject imaging on the base of those orthogonal functions, so it cancompletely avoid the bad influence due to nouniform baselines. BG inversealgorithm can also utilizes the priori information about the size of imagingscene to improve the height resolution.