Combined Centre-Weighted Median Filter and BM3D to Filter Digital Images in Mixed Gaussian and Impulsive Environments

Abstract

This paper studies the combination of a centre-weighted median filter with block matching 3-D filtering for digital images with mixed impulsive and Gaussian noise. The proposed combined filter has an extended application range with respect to both components. The work in the paper derives a simple formula for selecting the weight of the centre-weighted median filter. Adaptive weights are given for neighbourhoods 3 × 3 and wider as functions of the impulsive noise percentage and Gaussian noise strength. Simulation results confirm the effectiveness of the proposed combined filter.

KEYWORDS:

• BM3D filter
• Centre-weighted median filter
• Gaussian noise
• Image filtering
• Impulse noise

Additional information

Notes on contributors

Igor Djurović

Igor Djurović was born in Montenegro in 1971. He received BS, MS, and PhD degrees, all in electrical engineering, from the University of Montenegro, in 1994, 1996, and 2000, respectively. He is currently a professor with the University of Montenegro. He has been engaged in research activities in around 20 international research centres. He has published more than 200 papers in international scientific journals and conferences. He is co-author of a research monograph and author or co-author of six book chapters. He has also been a member of the editorial board of several international journals. From 2011, he has been an associated member of the Montenegrin Academy of Sciences and Arts. He was director of the first Montenegrin Centre of Excellence in Bio-informatics (BIO-ICT) and member and head of the research teams of numerous other projects. In 2016, he was awarded with the highest Montenegrin prize for scientific achievement. His current research interests are in the area of time-frequency analysis and applications for digital-image processing.

E-Mail: [email protected]