Airborne InSAR Assisted SAR Image Matching

SAR stereograms can extract high-precision three-dimensional surface information,combined with the sparse control points regional network adjustment,becoming a more robust SAR mapping method.Airborne SAR is easy to acquire interference and non-interference data,and it has controllable acquisition time and space.It has also been widely used in actual mapping.However,the joints of the same name needed for regional network adjustment of rare control points are basically selected manually,and the efficiency of mapping is low.The geometric distortion between airborne SAR images is serious,and the gray scale difference is large,resulting in low accuracy of matching between airborne SAR images.How to realize automation and high-precision matching has become a technical difficulty in the practical application of airborne SAR three-dimensional mapping.This paper aims at the matching problems mentioned above,and fully explores and utilizes auxiliary data of SAR images,mainly including InSAR and POS data,and proposes InSAR-assisted airborne SAR image matching method.The main research contents of this article include:1.InSAR elevation and SAR image geocoding:The traditional image geocoding relies on elevation data with high-precision geocoordinate information.The original InSAR phase elevation has elevation information but no geographic plane coordinate information.This article uses POS data.With InSAR phase elevation data,the InSAR data was first geocoded to obtain InSAR-encoded DSM products.The SAR image was then corrected using InSAR DSM with geographic coordinates.The focus is on the InSAR phase elevation direct method and indirect geocoding method supported by POS data.2.Airborne InSAR assisted SAR image matching:Due to the complex grayscale distribution of SAR images and the influence of speckle noise,the success rate and accuracy of SAR image matching has always been a difficult point in the automatic processing of SAR images.This paper proposes an InSAR phase elevation data-assisted SAR image matching method,which uses the same features of the same location of InSAR pixels and SAR image pixels,and the relative deformation of the encoded InSAR DSM in a certain area,through matching between InSAR DSMs.With the help of the strict geometric transformation relationship between InSAR and SAR data,the SAR image matching point of the same name is acquired indirectly,which provides a new attempt for SAR image matching and provides a new solution to the problem of SAR image auto-emptiness.3.Verification of InSAR-assisted SAR image matching technique:The matching point of the reference image is manually measured on the image to be matched,and the matching technique is used to obtain the point of the same name for comparison.The matching accuracy and matching accuracy of the SAR image are obtained;the comparison with the results of the classic SIFT matching method is performed to match the text.The effectiveness and performance of the method are evaluated.Theoretical methods and experiments show that the matching accuracy of the method given in this paper is about 70%,and the matching accuracy is 1 to 3 pixels.The accuracy and precision are both higher than the traditional airborne SAR image direct matching algorithm.