Impulsive noise filter using median- and partition-based operation

Abstract

A novel filter is proposed in this work to remove noise and preserve fine details from the impulse-corrupted images based on median and partition operation. Each pixel is classified into one of the mutually exclusive partition blocks by a novel observation vector. This observation vector can be directly used to partition the observation space without using predefined thresholds. The weight of each block is obtained using a learning approach based on the least-mean-square algorithm. The noise-filtering process is progressively applied through several iterations to achieve better noise-suppression results. Several experimental results have demonstrated that the proposed filter outperforms many filters in terms of both noise suppression and detail preservation.

Improved Radiometric Based Method for Suppressing Impulse Noise from Corrupted Images

“…That is to say c x and f x count these neighbor pixels, satisfying and un-satisfying the condition of |u y − u x | ≤ T : y ∈ Ω (N) i, j , respectively. Here, threshold T is set as T = 5 in the whole work, the same as the threshold d in [15]. Since there are (2N + 1) 2 − 1 neighbor positions in a filter window, the sum of c x and f x is (2N + 1)…”

“…Let c x and f x denote the number of "close" and "further" neighbor pixels which have the absolute gray differences to the center pixel u x not-exceeding and exceeding threshold T [15], respectively. That is to say c x and f x count these neighbor pixels, satisfying and un-satisfying the condition of |u y − u x | ≤ T : y ∈ Ω (N) i, j , respectively.…”