EM-based simultaneous registration, restoration, and interpolation of super-resolved images

Abstract

We present a maximum likelihood ( M l ) solution to the problem of obtaining high-resolution images from sequences of noisy, blurred, and low-resolution images. In our formulation, the registration parameters of the low-resolution images, the degrading blur, and noise variance are unknown. Our algorithm has the advantage that all unknown parameters are obtained simultaneously using all of the available data. An efficient implementation is presented in the j-equency domain, based on the Expectation Maximization (EM algorithm.Simulations demonstrate the eflectiseness of the algorithm.

A Blind Super-Resolution Reconstruction Method Considering Image Registration Errors

“…A few investigations have been carried out to estimate the HR image and blurring function simultaneously, to reduce the effect of blur estimation error. Authors in [23][24][25] proposed blind SR that can handle parametric blur models with one parameter. This restriction is, unfortunately, very limiting for most real applications.…”

Compstat 2006 - Proceedings in Computational Statistics

“…Authors in [NMG01, WGK03] proposed BSR that can handle parametric PSFs, i.e., PSFs modeled with one parameter. A good survey is for example in [PPK03].…”

Estimation of High Resolution Images and Registration Parameters from Low Resolution Observations

“…Most of the reported approaches on shift estimation for super-resolution, first estimate the displacement vectors either by interpolating the low resolution observations and then finding the registration parameters or by finding the low resolution registration parameters in the low resolution domain and then interpolating them (consider [4] again), where the high resolution image and the registration parameters are estimated simultaneously, can be found in [10,9,12,2].…”