Noise suppression and barrier crossing in Monte Carlo image-restoration method

摘要

Abstract: In this paper, an efficient approach for imagerestoration of noisy data is suggested. This approachcombines the Monte Carlo image restoration techniqueand the Morrison noise removal methods. The meansquared error (MSE) criterion is used to test theperformance of the Monte Carlo method with and withoutprior-application of the Morrison noise removal method.The methods for facilitating the Monte Carlo walk tothe brightest regions of the image are discussed and anew approach is suggested. It is shown that the MonteCarlo technique is potentially very fast with goodresolution. The Morrison noise removal method smoothesthe data at the first iteration and proceeds to restorethe data back to its original noisy form at lateriterations. To achieve some noise suppression, one canstop the Morrison iterations before it converges to theoriginal noisy form. The Monte Carlo method is thenapplied to the noise suppressed data. !19