A multi-images variational Bayesian super-resolution reconstruction method based dual sparse priors

Abstract

Aiming at solving the problem of prior constraints on variational bayesian super-resolution reconstruction method, we propose a novel prior model to overcome the under-constraint of non-edge regions of image due to total variation prior, so the generation and spread of noise are further suppressed. We combine the weighted total variation model and L1 norm model, achieving a variational bayesian super-resolution reconstruction method based dual sparse priors. The super-resolution results of the simulation data and real data demonstrate that our algorithm is more effective and stable than the same type of other methods.