The Efficaciousness of the Outlier Reclamation Scheme Established on a First-Order Recursive FIR Linear Predictor for Salt&Pepper Outlier
Vorapoj Patanavijit, Darun Kesrarat, Wilaiporn Lee, Kanabadee Srisomboon, Kornkamol Thakulsukanant · 2024
Typically, the concomitant outlier reclamation scheme, which is ideologically integrated of an outlier unmasking scheme and an outlier reclamation scheme, is an obligatory pre-computing scheme that has to be operated before other sophisticated image processing schemes. As a result, a First-Order Recursive Linear Finite Impulse Response Linear Predictor (FRL-FIRLP) that has been evolved from a pioneer median filter (SMF) has been asserted for Salt and Pepper noise at great propagation. This technical essay addresses to interrogate the efficaciousness and limitation of a FRL-FIRLP scheme, which offers the statistical concept of reclamation of the noisy component by using a linear predictor established on ARMA (Auto Regressive Moving Average). In the computer verification section, countless noisy decorations, which are integrated of Lena and Girl at great propagation, are used to verify the efficaciousness of the outlier reclamation scheme established on FRL-FIRLP scheme. A correlation of the efficaciousness is made with countless current schemes (Median Filter (SMF), Mean Filter (MF), Adaptive Median Filter (AMF)) in terms of visional and measurable aftermaths (PSNR and SSIM).