The Research On Image Registration Of InSAR

Interferometric Synthetic Aperture Radar (InSAR) image technique is a new radar image technique developed from Synthetic Aperture Radar (SAR).Carried forward the merit of SAR, InSAR can create high resolution DEM and accomplish change detection, such as slow-moving targets or subtle shift of the earth's surface. It can be applied to many areas of military and civil.First the paper introduces the trait and its extra data format of SAR. It researches the new method for data processing.During the data processing of InSAR,for we get not the images but the raw data of InSAR. Registration of two or more images of the same scene is an important procedure in InSAR image processing that seeks to extract differentiao phase information not obtainable from each one of these images. Meanwhile, the accuracy of this step is crucial to the reliability of subsequent image processing and final resules of the data processing chain, meeting subpixel precision request. So it is necessary to research the images registration before unwrapping.The paper anaysizes some conventional InSAR registration methods, based on the characteristics of InSAR image the paper presents two multi-step image matching algorithms.One method is integrating correlation-registration and neural network image registration, the other is integrating correlation-registration and spectral maxiumun registration. In coarse registration, the two methods all utilize correlation-registration to obtain pixel accuracy registration results, offering crude control points pairs for the precise matching. In precise image registration, two methods respectively apply neural network and spectral maxiumun to register images, the results of experiment show that the methods was resultful.