The Registration Algorithms Research Of InSAR Image

Interferometric Synthetic Aperture Radar(InSAR) is a technique that derives the elevation data and slight deformation of earth surface by extracting phase information as the data source in complex information images of Synthetic Aperture Radar.The registration of two or more images of the same area is an important procedure in the entire InSAR image processing.The matching pixel-accuracy of two images is important to the quality of interferogram phase.For more accurate interferogram phase,the two images registration need to be the sub-pixel precision level.The basic principle and devoloping status of InSAR imaging technique is introduced in the thesis.Also,we make the intensive study of the methods of InSAR complex images registration,especially the two sub-pixel registration algorithms.One is the correlation matching algorithm,the other is the spectral maximum matching algorithm.Then we do some registration experiment on the SLC data of KeShi and WuluMuQi.By the interferogram phase and the correlative data analyzing,the experiment effect proves to be the satisfied result.Then the common interpolation and resampling methods in slave image of InSAR were discussed and compared.In the end of the paper,on the basis of analyzing the characteristics of single complex data(slc)and format of raw data of insar,we do some research on reading slc data and do some registration experiment of KeShi panoramic view.