A general sparse image prior combination in Compressed Sensing

In this paper a general combination of sparse image priors is applied to Bayesian Compressed Sensing (CS) reconstruction of digital images. A simultaneous deblurring and CS reconstruction variational algorithm is derived. The application of the new algorithm, to both blurred and non-blurred images at different compression ratios, is studied. The new method is applied to Passive Millimeter-Wave Imaging (PMWI) CS. and its performance compared to state of the art CS reconstruction methods.