Study And Engineering Implementation On Problem Areas Detection Technology For Spaceborne InSRA

Interferometric Synthetic Aperture Radar(In SAR)is able to quickly acquire Digital Elevation Model(DEM)with high accuracy through interferometric processing of multiple SAR images.Spaceborne In SAR system has obtained domestic and international attention.However,it also faces many problems.Because of the complexity of terrain undulations,shadow and layover will appear within the SAR images.In addition,along with the existence of water areas,phase unwrapping would be so difficult in large areas that even end up in failure,which seriously affects the accuracy of DEM.In this paper,shadow,layover,and water areas are called problem areas.In order to robustly detect the problem areas in In SAR processing,this paper has carried out the following works:(1)Analyze the formation process of shadow and layover,as well as the characteristics of intensity and interference of shadow,water,and layover on the basis of the imaging principle.(2)Propose a coarse-to-fine threshold estimation method to detect the problem areas.Firstly,the coarse threshold interval of the problem areas is obtained based on the Otsu multiple thresholds technique.Then,an iterative Otsu threshold method is proposed,which can achieve the optimal threshold through maximizing the normalized inter-variance.The proposed iterative Otsu threshold is utilized to further determine the intensity threshold of the problem areas from the coarse threshold interval.Because this method is only dependent on intensity information,it has higher calculation efficiency.The experiments on the real data show that the proposed method is able to effectively improve the accuracy and robustness for the problem areas detection.(3)Propose a novel method for detecting problem areas based on the combination of multi-source information.The proposed method efficiently takes advantage of the prior DEM information,intensity information of SAR images,In SAR coherence coefficients to detect shadow,and applies Global Land Water Mask Data,intensity information of SAR images,and phase information to detect layover.The experiments on real data show that the proposed method can effectively solve the inaccuracy problem when using only one source information.(4)Perform engineering application of the proposed method for detecting problem areas based on the combination of multiple-source information.As a result,the problem areas can be detected automatically.