Innovative approach on novel open-close sequence filters to remove impulse noise in highly corrupted images

Abstract

In this research paper, various impulse noise filters are studied. These are median filter, center weighted median filter, pixel-wise MAD (mean of the absolute deviations from the median) filter and. Here a novel, effective and simple approach is implementing using powerful open-close sequence (OCS) filter to restore images that are corrupted by 40%–80% probability impulse noise. Extensive simulation results show that proposed method substantially outperforms all other existing method, in terms of suppressing impulse noise while preserving image details. Since proposed method is algorithmically simple and requires low computational complexity, it is suitable to be implemented by hardware and applied to many real-time applications