INCORPORATION OF SPATIAL INFORMATION IN BAYESIAN IMAGE-RECONSTRUCTION - THE MAXIMUM RESIDUAL LIKELIHOOD CRITERION

We have developed a new figure of merit, a "maximum-residual-likelihood" (MRL) statistic, for the goodness of fit for Bayesian image restoration which explicitly incorporates spatial information. The MRL constraint provides a natural means of incorporating the prior knowledge that the residuals contain no spatial structure through the autocorrelation function of the residuals. We demonstrate that this statistic follows a chi2 distribution and that forcing this statistic to have its most probable value leads to a restored image whose residuals are consistent with the noise model. Our numerical experiments suggest that image restoration using the MRL statistic alone (without an "image prior," e.g., an entropy function) is numerically robust and produces results which are independent of the initial guess for the restored image. However, we caution that using the MRL statistic without an image prior can result in overresolution in low signal-to-noise portions of the image.