Sentinel-1A/B data are crucial for retrieving numerical information about surface phenomena and processes.Coregistration of terrain observation by progressive scans(TOPS)data is a critical step in its application.TOPS...Sentinel-1A/B data are crucial for retrieving numerical information about surface phenomena and processes.Coregistration of terrain observation by progressive scans(TOPS)data is a critical step in its application.TOPS data must be fundamentally co-registered with an accuracy of 0.001 pixels.However,various decorrelation factors due to natural vegetation and seasonal effects affect the coregistration accuracy of TOPS data.This paper proposed an enhanced spectral diversity coregistration method for dual-polarimetric(PolESD)Sentinel-1A/B TOPS data.The PolESD method suppresses speckle noise based on a unified non-local framework in dual-pol Synthetic Aperture Radar(SAR),and extracts the phase of the optimal polarization channel from the denoised polarimetric interferometric coherency matrix.Compared with the traditional ESD method developed for single-polarization data,the PolESD method can obtain more accurate coherence and phase and get more pixels for azimuth-offset estimation.In bare areas covered with low vegetation,the number of pixels selected by PolESD is more than the Boxcar method.It can also correct misregistration more effectively and eliminate phase jumps in the burst edge.Therefore,PolESD will help improve the application of TOPS data in low-coherence scenarios.展开更多
We propose a multiscale projection method for the numerical solution of the irtatively regularized Gauss-Newton method of nonlinear integral equations.An a posteriori rule is suggested to choose the stopping index of ...We propose a multiscale projection method for the numerical solution of the irtatively regularized Gauss-Newton method of nonlinear integral equations.An a posteriori rule is suggested to choose the stopping index of iteration and the rates of convergence are also derived under the Lipschitz condition.Numerical results are presented to demonstrate the efficiency and accuracy of the proposed method.展开更多
基金supported by Jilin Changbaishan Volcano National Observation and Research Station(Project No.NORSCBS20-04)National Natural Science Foundation of China(42174023)the Fundamental Research Fund for the Central Universities of Central South University(No.506021722).
文摘Sentinel-1A/B data are crucial for retrieving numerical information about surface phenomena and processes.Coregistration of terrain observation by progressive scans(TOPS)data is a critical step in its application.TOPS data must be fundamentally co-registered with an accuracy of 0.001 pixels.However,various decorrelation factors due to natural vegetation and seasonal effects affect the coregistration accuracy of TOPS data.This paper proposed an enhanced spectral diversity coregistration method for dual-polarimetric(PolESD)Sentinel-1A/B TOPS data.The PolESD method suppresses speckle noise based on a unified non-local framework in dual-pol Synthetic Aperture Radar(SAR),and extracts the phase of the optimal polarization channel from the denoised polarimetric interferometric coherency matrix.Compared with the traditional ESD method developed for single-polarization data,the PolESD method can obtain more accurate coherence and phase and get more pixels for azimuth-offset estimation.In bare areas covered with low vegetation,the number of pixels selected by PolESD is more than the Boxcar method.It can also correct misregistration more effectively and eliminate phase jumps in the burst edge.Therefore,PolESD will help improve the application of TOPS data in low-coherence scenarios.
基金Supported in part by the Natural Science Foundation of China under grants 11761010 and 61863001.
文摘We propose a multiscale projection method for the numerical solution of the irtatively regularized Gauss-Newton method of nonlinear integral equations.An a posteriori rule is suggested to choose the stopping index of iteration and the rates of convergence are also derived under the Lipschitz condition.Numerical results are presented to demonstrate the efficiency and accuracy of the proposed method.
