A Circular Adaptive Median Filter for Salt and Pepper Noise Suppression from MRI Images

Abstract

An adaptive median filter with a circular kernel, named as circular adaptive median filter (CAMF) is proposed in this article, for denoising of magnetic resonance imaging (MRI) images corrupted by salt and pepper noise of varying noise densities. An adaptive operation is incorporated in the proposed filter by varying the size of the circular kernel according to the requirement. The effectiveness of the CAMF is compared with six other competitive networks, i.e., conic adaptive median filter (CoAMF), decision based filter (DBF), modified switching median filter (MSWM), recursive adaptive modified filter (RAMF), plus adaptive median filter (PAMF), and cross adaptive median filter (CrAMF). The performance of all the models is analyzed using peak signal to noise ratio (PSNR) and computational time. Moreover, a non-parametric statistical test is conducted to illustrate the pair wise comparison of other filter with the proposed one. It is observed that the proposed approach has demonstrated superior performance with respect to the two performance measures.