PROJECT TITLE :

Generalized Bayesian Model Selection for Speckle on Remote Sensing Images

ABSTRACT:

Coherent summation of back-scattered waves and subsequent nonlinear envelope changes introduce speckle noise into both synthetic aperture radar (SAR) and ultrasound (US), two major active imaging techniques for remote sensing. In order to create denoising methods and improve statistical inference from remote sensing images, it is essential to know the features of this multiplicative noise A recently devised RJMCMC-based Bayesian technique, trans-space RJMCMC, has been applied in this paper with the reversible jump Markov chain Monte Carlo algorithm. SAR and US remote sensing images may be automatically classified into model classes based on popular envelope distribution families, according to the proposed method. To avoid an exhaustive search, the suggested method calculates the right distribution family as well as the shape and scale parameters. Various SAR photos of urban, forest, and agricultural landscapes, as well as two separate US images of a human heart, were employed in the experimental study. Speckle statistical models can be found using the proposed method, as demonstrated by simulation results.