Performance Analysis and Optimization of Nonlinear Image Restoration Techniques in Spatial Domain

This paper is concerned with critical performance analysis of spatial nonlinear restorationtechniques for continuous tone images from various fields (Medical images, Naturalimages, and others images).The performance of the nonlinear restoration methods isprovided with possible combination of various additive noises and images from diversifiedfields. Efficiency of nonlinear restoration techniques according to difference distortion andcorrelation distortion metrics is computed. Tests performed on monochrome images, withvarious synthetic and real-life degradations, without and with noise, in single framescenarios, showed good results, both in subjective terms and in terms of the increase ofsignal to noise ratio (ISNR) measure. The comparison of the present approach withprevious individual methods in terms of mean square error, peak signal-to-noise ratio, andnormalized absolute error is also provided. In comparisons with other state of art methods,our approach yields better to optimization, and shows to be applicable to a much widerrange of noises. We discuss how experimental results are useful to guide to select theeffective combination. Promising performance analyzed through computer simulation andcompared to give critical analysis.