The Research On Interferogram Denoising Algorithm Of InSAR

Interferometric synthetic aperture radar (InSAR) imaging technique has many advantages over other imaging methods, such as large coverage, high spatial resolution, high elevation accuracy, and the capability of working at all time and under all-weather conditions. It is an effective method to acquire the three-dimensional Digital Elevation Model (DEM). So it can be wilely used for topographic surveying, earthquake monitoring, volcano monitoring, deformation detection and so on.By summarizing the development history and the present situation of InSAR three-dimensional imaging technique, analyzing the principle of InSAR system and main factors which affect the InSAR measuring accuracy, the data processing flow of InSAR system is presented. According to the difference of interference phase from two SAR plural images, the InSAR technique gains the three-dimensional information of ground object, therefore the noise of interferogram is one of the most important factors which affect the measuring accuracy of the system. In order to obtain accurate data of interference phase to make two-dimensional phase unwrapping done, we must denoise effectively. So the research on Interferogram denoising algorithm of InSAR is needed.By researching the origin, characteristic of the InSAR interferogram noise and statistical characteristic of phase, the advantage and disadvantage of traditional interferogram filtering methods are analyzed based on experiment results. By involving the wavelet analysis into interferogram denoising, some research work about interferogram denoising method of InSAR based on wavelet transform and periodic mean/median algorithm is accomplished and an interferogram denoising method based on wavelet transform and weighted periodic pivoting median filter algorithm is proposed. In the synthesis consideration on the statistical characteristic and directional characteristic of high frequency coefficient, an interferogram denoising method based on wavelet transform and directional median is proposed. The method is built on wavelet transform, then checks the high frequency coefficient of all the orientations, judges whether the coefficient is the edge of corresponding orientation or not, smoothes the coefficient using different mode corresponding to its orientation and filters the coefficient using median filtering. Experiments have been carried out based on ERS-1/2 SAR tandem data and the results show that the proposed methods can reduce interferogram noise efficiently as well as preserve edge and detail very well. The methods also guarant the interference fringe smooth and establish the foundation for the later data processing and research work.