Accurate Despeckling and Estimation of Polarimetric Features by Means of a Spatial Decorrelation of the Noise in Complex PolSAR Data

Abstract

In this work, we extended a procedure for the spatial decorrelation of fully-developed speckle, originally developed for single-polarization SAR data, to fully-polarimetric SAR data. The spatial correlation of the noise depends on the tapering window in the Fourier domain used by the SAR processor to avoid defocusing of targets caused by Gibbs effects. Since each polarimetric channel is focused independently of the others, the noise-whitening procedure can be performed applying the decorrelation stage to each channel separately. Equivalently, the noise-whitening stage is applied to each element of the scattering matrix before any multilooking operation, either coherent or not, is performed. In order to evaluate the impact of a spatial decorrelation of the noise on the performance of polarimetric despeckling filters, we make use of simulated PolSAR data, having user-defined polarimetric features. We optionally introduce a spatial correlation of the noise in the simulated complex data by means of a 2D separable Hamming window in the Fourier domain. Then, we remove such a correlation by using the whitening procedure and compare the accuracy of both despeckling and polarimetric features estimation for the three following cases: uncorrelated, correlated, and decorrelated images. Simulation results showed a steady improvement of performance scores, most notably the equivalent number of looks (ENL), which increased after decorrelation and closely attained the value of the uncorrelated case. Besides ENL, the benefits of the noise decorrelation hold also for polarimetric features, whose estimation accuracy is diminished by the correlation. Also, the trends of simulations were confirmed by qualitative results of experiments carried out on a true Radarsat-2 image.

A No-Reference Edge-Preservation Assessment Index for SAR Image Filters under a Bayesian Framework Based on the Ratio Gradient

“…In particular, during the SAR focusing operation, the matched filters are used. The above processes can introduce a spatial correlation of the noise, which may impact the validity of IID assumption [30]. Therefore, in theory, an initial decorrelation step is better to be taken on the speckled image, before applying the proposed IID assumption based index.…”

“…Instead, the post-processing approaches are better choices. Among them, the downsampling strategy and the method presented by Arienzo et al [30] are two typical approaches. The downsampling strategy can directly reduce the spatial correlations of speckle by scarifying the resolution of SAR images.…”

“…The downsampling strategy can directly reduce the spatial correlations of speckle by scarifying the resolution of SAR images. In [30], a blind decorrelation method is proposed, of which the main idea is to estimate and invert the point spread function by performing least square optimization. However, this blind decorrelation method can only handle the decorrelation issue of SLC SAR data, which limits its applicability before calculating the RGPI values As aforementioned, to weaken the influence of speckle, the ratio gradient in the RGPI is calculated using the mean values of two square patches contiguous to a given pixel, instead of using the values of two single pixels.…”

“…Since the RGPI is derived based on IID assumption, it is necessary to further inspect its validity on more images with correlated speckle. To this end, we used the clean "building" image provided in [14] and generated the speckle correlated single-look images by the approach presented in [30]. In this simulation process, the choice of parameter α which determines the degree of spectral smoothness is crucial.…”

Multi-Objective CNN-Based Algorithm for SAR Despeckling

“…As reported in the literature, these points should be left unfiltered or at least processed in a different way [44]. Some methods, such as SAR-BM3D, NOLAND, and fast adaptive nonlocal SAR despeckling (FANS), filter them by aggregation of similar patches selected through a statistical approach; other methods, such as [45], do not filter them at all.…”