Removal of high-density impulse noise based on switching morphology-mean filter

Abstract

In this paper, an effective switching morphology-mean filter is proposed to remove impulse noise from digital images. The proposed method consists of two stages. The first stage is to detect the impulse noise in an image. The second stage is to construct adaptive structuring elements and remove the impulse noise. In this stage, the “noise-pixels” are filtered out by using the morphology-mean filter, and the “noise-free pixels” are copied directly to the output image. The proposed filter uses the window of size 3×3pixels. Simulation results show that the proposed method outperforms some existing algorithms for high-density impulse noise, both in vision and quantitative measurements. Moreover, our method can preserve effectively the details in an image, and has less processing time.