A Pervasive Achievement of Iterative Mean Filter for An Outlier Extinguish Objective on FVIN

Abstract

For outlier extinguish investigating population, especially on the impulsive outlier, the outlier extinguish approach is a prevalent process, which is first executed before other progress operations therefore the iterative mean filter (IMF) is originated for FVIN (Fix Value Impulsive Noise) and turned into one of the most achievement outlier extinguish approach. This academic article addresses to prospect the outlier extinguish approach established on IMF, which is originated from mean filter in order to carry out the deficiency of the foregoing outlier extinguish approach, for FVIN at omnipresent scattering of outlier consistency (5%-90%). In the experimental analysis, abundant outlier figures, which are unified of F16, Girl, Lena and Pepper at omnipresent scattering of outlier consistency, are examined for analyzing the achievement of an outlier extinguish approach established on iterative mean filter. From these results, the outlier extinguish approach established on IMF has superior achievement, in contrast with SMF, MF, AMF.