Search

Synthetic aperture radar is an efficient remote sensing tool by producing high spacial resolution images. But, synthetic aperture radar data suffer speckle noise effect that difficult their processing (for example, making boundary detection). We propose and assess edge detectors for synthetic aperture radar imagery based on stochastic distances between models. These edge detectors stem from generalized divergences with good asymptotic properties. Results reveal that divergence-based detectors can outperform the likelihood-based counterpart. [ABSTRACT FROM AUTHOR]

The proposal of more exible distributions is an activity often required in practical contexts. In particular, adding a positive real parameter to a probability distribution by exponentiation of its cumulative distribution function has provided exible generated distributions having interesting statistical properties. In this paper, we study general mathematical properties of a new generator of continuous distributions with three extra parameters called the exponentiated Gompertz generated (EGG) family. We present some of its special models as well as an essay on its physical motivation. From mathematical point of view, we derive explicit expressions of the EGG family: the ordinary and incomplete moments, quantile and generating functions, Bonferroni and Lorenz curves, Shannon and Rényi entropies and order statistics, which are valid for any baseline model. We also provide a bivariate EGG extension. The estimation procedure by maximum likelihood of the new class is elaborated and discussed. In order to quantify and to assess the asymptotic behavior of this procedure, we perform a simulation study. Finally, two applications to real data are performed. Results furnish evidence in favor of the use of the EGG beta distribution as a good proposal to these data sets. [ABSTRACT FROM AUTHOR]

This work presents a comprehensive examination of the use of information theory for understanding polarimetric synthetic aperture radar (PolSAR) images by means of contrast measures that can be used as test statistics. Due to the phenomenon called 'speckle', common to all images obtained with coherent illumination such as PolSAR imagery, accurate modelling is required in their processing and analysis. The scaled multilook complex Wishart distribution has proven to be a successful approach for modelling radar backscatter from forest and pasture areas. Classification, segmentation, and image analysis techniques, which depend on this model, have been devised, and many of them employ some kind of dissimilarity measure. Specifically, we introduce statistical tests for analyzing contrast in such images. These tests are based on the Bhattacharyya, chi-square, Hellinger, Kullback-Leibler and Rényi distances. Results obtained by Monte Carlo experiments reveal the Kullback-Leibler distance as the best one with respect to the empirical test sizes under several situations which include pure and contaminated data. The proposed methodology is applied to actual data, obtained by an E-SAR sensor over surroundings of Weßling, Bavaria, Germany. [ABSTRACT FROM AUTHOR]