SAR interferometry: phase unwrapping by fringe-line detection

Abstract

ABSTRACT Two-dimensional phase unwrapping is the key technique in SAR interferometry. This paper discusses a phaseunwrapping algorithm based on fringe line detection. Improvements are presented in order to use this method in areas with intensely fluctuated fringe lines. Results on X-SAR interferometric data are obtained. They confirm the excellent performance ofthe procedure.key words: SAR, interferometry, phase unwrapping, fringe line detection 1. INTRODUCTION Interferometric SAR (INSAR) is a field of growing interest. By exploiting the interferogram computed from a SARdata pair acquired by two spatially separated side views of the same site, high-resolution terrain digital elevation models(DEMs) could be generated.There are four main steps in SAR interferometry: image coregistration, phase interferometry, two-dimensional phaseunwrapping and terrain elevation determination. As the accuracy of the terrain elevation mapping with INSAR dependscrucially on the phase unwrapping, the phase unwrapping algorithm becomes the key technique in SAR interferometry.Phase jump and phase undersamplmg are two main problems' in phase unwrapping. Phase jump is caused by ameasurement limitation and discontinuity lines are generated at each 2ir excursion of phase. Phase undersampling usuallytaking place in layover conditions is caused by a time discretization limitation when the measured phase is sampled at a rateinsufficient for its faithful reconstruction. It will be impossible to solve the latter problem without extra informationconcerning the lost 2Nit factor.Some practical algorithms have been proposed, such as the classic phase unwrapping method,' phase unwrapping byKalinan filtering,2 phase unwrapping through estimating the phase derivative in two orthogonal directions,3 phaseunwrapping using Green's formulation,4 phase unwrapping in the least square sense,5 etc.A novel phase unwrapping technique by fringe line detection6 was proposed by Qianlin in 1992. This algorithm isespecially suitable for images with large phase noise, which results in large number of residue regions.