| The rapid development of radar remote sensing technology provides broad and rapid scene information for surface water investigation and monitoring.Among them,Polarimetric Synthetic Aperture Radar(Pol SAR)has abundant scattering information of ground objects compared with the traditional SAR system.Meanwhile,the dual-polarized SAR system has a larger imaging width than the full-polarized SAR system.The more convenient way of data acquisition is more conducive to the large-scale surface water remote sensing research.However,most of the existing studies on the spatial distribution of water bodies are based on single-polarization SAR data,and the potential of dual-polarization information is rarely explored.Since traditional SAR data does not have polarization scattering information,in the past,auxiliary information such as DEM data was often used to distinguish some weakly scattered objects with similar backscattering coefficient values to water target in SAR images,such as shadow,road and bare soil.However,DEM data with high accuracy was often difficult to obtain.At the same time,in the aspect of monitoring suspended solids in surface water research,in the past,more multi-spectral and hyperspectral data were used to study suspended solids based on its performance of scattering and absorption of light radiation,but the potential of geometric monitoring of suspended solids surface by Pol SAR microwave band was ignored to improve the accuracy of quantitative estimation and monitor the spatial and temporal distribution of suspended solids.In view of the above problems,the main objective of this study is to explore the potential of dual-polarization polarization information in the fine extraction of surface water information.The main contents and innovations are as follows.1.Applicability analysis of dual-polarization data of common SAR and fully polarized SAR water extraction methods.In view of the current situation that most of the commonly used water information extraction methods are applied to single-polarization SAR data,and a small number of fully polarized SAR data are used to classify ground object scenes including water objects,this study compared and analyzed the applicability of these methods in different water environment types of dual-polarization data.There are different surrounding environments such as mountains,flat areas and cities.2.A method to distinguish water objects from easily confused non-water objects based on polarization entropy difference((35)H)is proposed.Aiming at the problem of distinguishing between water and easily confused non-water weak scattering targets,the existing dual-polarization characteristics of these targets were analyzed to explore the effective features.Then,combining with the environmental factors of the target,a new dual-polarization feature--polarization entropy difference((35)H)was constructed.Experiments show that the adaptive threshold method can be used to effectively distinguish between water and easily confused non-water targets by using this feature.3.A method of extracting suspended solids from water body based on improved anisotropy parameter(A’)was proposed.Since the calculation process of dual-polarization target decomposition(H/ α decomposition)is complicated,and the physical meaning of the expression is not intuitive.Later,get a new parameter by improving the A parameter.Compared with the polarization entropy(H),the physical meaning expressed by this parameter is more intuitive.And it also has a certain speed improvement for the parameter calculation of big data.Subsequently A’ is used to construct a new classification plane A/’ α similar to H/ α plane,which can also effectively classify targets.Combined with A/’ α plane,by analyzing the features of suspended solids area and clear water area,effective features are selected for further classification of internal water area.The classification results show that this method is suitable for monitoring the spatial distribution of suspended solids. |
PDF Full Text Request