A Novel Robust Outlier Classification Founded On Localised Rank-Ordered Logarithmic Differences For Fix-Value Impulsive Noise

Outlier classification founded on Localised Rank Ordered Difference (LROD) technique, which is developed from Rank-Ordered Absolute Differences (ROAD), has been presented since 2015 therefore this LROD technique becomes one of the ultimate efficient outlier classification from its performance. This paper proposes the novel and robust outlier classification founded On LROLD (Localised Rank-Ordered Logarithmic Differences) technique, which is developed from LROD and Rank-Ordered Logarithmic Differences (ROLD), which is more efficient than LROD, for Fix-Value Impulsive Noise (FVIN). From the experimental simulation on three testing portraits: Girl, Pepper and Lena, the robust outlier classification founded on LROLD technique has the better efficiency than the previous techniques: LROD technique and ROAD technique at plentiful distribution of FVIN.