Font Size: a A A

Research On Multibaseline InSAR Phase Unwrapping Algorithm Based On Cluster-Analysis

Posted on:2023-06-10Degree:MasterType:Thesis
Country:ChinaCandidate:T J ChenFull Text:PDF
GTID:2558306911956909Subject:Electronic Science and Technology
Abstract/Summary:PDF Full Text Request
Interferometric synthetic aperture radar(InSAR)can obtain target elevation information,which has the characteristics of all-weather and all-weather.Among them,phase unwrapping technology plays an important role.At first,researchers used single baseline phase unwrapping(SBPU)technology to recover elevation information.In order to overcome the shortcomings of SBPU,researchers further proposed multi baseline phase unwrapping technology.Through the mutual combination of multiple baselines,the change of terrain elevation corresponding to the interference phase change is improved,the phase continuity hypothesis is overcome,and the accuracy of phase unwrapping is improved under the adverse effects of terrain sudden change,huge external noise and overlapping effect.In this paper,the multi baseline InSAR phase unwrapping algorithm based on cluster analysis is studied.The main work includes:(1)A cluster correction method of MBPU algorithm based on cluster analysis is proposed.The first method takes the cluster number that appears most frequently in the selected window as the final cluster number of the central pixel of the window directly according to the maximum likelihood criterion;The second method first divides all pixels into core pixels and noncore pixels according to the size of density,and then changes the cluster number of noncore pixels to the cluster number with the most occurrences in the selected window according to the maximum likelihood criterion.(2)A BM3D denoising method of MBPU method based on cluster analysis is proposed.BM3D filtering technology is applied to the following four links of clustering algorithm:(1)BM3D filtering of wrapping interferogram;(2)BM3D filtering of intercept;(3)BM3D filter the cluster number obtained after clustering;(4)BM3D filtering is performed on the unwrapping phase.As mentioned above,this paper has made some achievements in the research of MBPU algorithm based on cluster analysis.In the aspect of MBPU based on cluster analysis,two methods to improve the accuracy of phase unwrapping are proposed.One is a kind of correction method,which modifies the wrong clustering result into the correct clustering result;The second is the filtering method,which applies BM3D filtering technology to the four links of cluster analysis.Both ideas will further improve the unwrapping accuracy.
Keywords/Search Tags:Interferometric Synthetic Aperture Radar, Multibaseline InSAR, Phase Unwrapping, Cluster Analysis, Cluster Correction, BM3D
PDF Full Text Request
Related items