Um novo algoritmo para filtragem de speckle em imagens SAR de intensidade baseado em distâncias estocásticas

Abstract

This dissertation work presents a new approach for filter design based on stochastic distances and tests between distributions. A window is defined around each pixel, overlapping samples are compared and only those which pass a goodness-of-fit test are used to compute the filtered value. Modified Nagao-Matsuyamawindows are used to define the samples. The technique is applied to intensity Synthetic Aperture Radar (SAR) data with homogeneous regions, using the Gamma model ¡(L,L/λ) with varying number of looks L allowing, thus, changes in heterogeneity. The proposal is compared with the Lee’s filter which is considered a standard, using a protocol based on Monte Carlo. We also analyzed the behavior of the filter on real data. Among the criteria used to quantify the quality of filters, we employ the equivalent number of looks (related to the signal-to-noise ratio), line contrast, and edge preservation. Moreover, we also assessed the filters by the Universal Image Quality Index and, the Pearson’s correlation on edges regions. Thesemeasuresmade use of the proposal on simulated data, which showed excellent results in the scenarios adopted. In particular, the proposed filter reduces the speckled, has preserved edges and texture features that keeps a good signal-to-noise in SAR data. Also been discussed applications in real SAR images, which were validated with numericalmeasurements. Method is quite general and can be extended to other conditions, provided that there is an underlying statisticalmodel. Thus, the new filter has been extended to operate in PolSAR images modeled by Wishart distribution Z ~W(Σ,L), where L is equivalent number of looks and Σ is a covariancematrix.