Abstract
In this paper, a novel two-phase modified decision based unsymmetrical trimmed mean filter, for removal of very high density salt and pepper noise (SPN) from images and videos is proposed. The first phase comprises the use of an unsymmetrical trimmed mean filter when the processing window is fully noisy and contains both outliers (0 and 255). The second phase is applied to eradicate the residual noisy pixels by replacing the processing pixel conditionally either with mean or with unsymmetrical trimmed mean. A second version of the algorithm is also devised by just replacing the unsymmetrical trimmed mean with unsymmetrical Winsorized mean in the second phase. The efficacy of different algorithms are evaluated, upto 99% density of SPN against standard Grey scale images, color images and video databases, in terms of Peak Signal to Noise Ratio (PSNR), Image Enhancement Factor (IEF) and Structural Similarity Index (SSIM). It has been observed that the proposed algorithms exhibit excellent noise suppression capabilities by giving high value of PSNR, IEF and SSIM. The edge preserving capability of the proposed algorithms is evaluated by Pratt’s Figure of Merit (PFOM), quantitatively and qualitatively it has been proved that edge preservation capability of the proposed algorithms is best among the state-of-art-algorithms.
Similar content being viewed by others
References
Abdou IE, Pratt WK (1979) . Proc IEEE 67(5):753
Aiswarya K, Jayaraj V, Ebenezer D (2010) .. In: IEEE Second International Conference on Computer Modeling and Simulation, vol 4, pp 409–413
Astola J, Kuosmanen P (1997) Fundamentals of nonlinear digital filtering, vol 8 (CRC press)
Bovik A (1987) . IEEE Transactions on Acoustics, Speech, and Signal Processing 35(4):493
Bovik AC (2005) Handbook of image and video processing (communications, networking and multimedia) (academic press inc.
Camarena JG, Gregori V, Morillas S, Sapena A (2008) . J Vis Commun Image Represent 19(1):20
Camarena JG, Gregori V, Morillas S, Sapena A (2010) . Image Vis Comput 28(1):188
Christo MS, Vasanth K, Varatharajan R (2020) . Multimedia Tools and Applications 79(1):415
Dash S, Das S, Panigrahi B (2017) International Conference on Intelligent Computing and Applications. ICICA. Advances in Intelligent Systems and Computing (Springer Singapore)
Esakkirajan S, Veerakumar T, Subramanyam AN, PremChand C (2011) . IEEE Signal Process Lett 18(5):287
Goel N, Aggarwal P (2019) Iranian journal of science and technology. Transactions of Electrical Engineering 43(3):459
Goel N, Kaur H, Saxena J (2020) Multimedia tools and applications (2020)
Hwang H, Haddad RA (1995) . IEEE Trans Image Process 4 (4):499
Jayaraj V, Ebenezer D (2010) . EURASIP Journal on Advances in Signal Processing 2010(1):690218
MATLAB (2020) version 9.8.0.1396136 (R2020a) (The MathWorks Inc., Natick Massachusetts)
Pratt WK (1978) Digital image processing. Wiley-Interscience, New York
Raza MT, Sawant S (2012) .. In: Nirma University International Conference on Engineering (NUiCONE) (IEEE, 2012), pp 1–5
Samantaray AK, Mallick P (2015) . Procedia Computer Science 48:222
Srinivasan K, Ebenezer D (2007) . IEEE Signal Processing Letters 14(3):189
Toh KKV, Isa NAM (2010) . IEEE Trans Consum Electron 56 (4):2560
(1977) The USC-SIPI Image Database. http://sipi.usc.edu/database/
Varatharajan R et al (2020) Computer Communications 2020
Vasanth K, Manjunath T, Raj SN (2015) . Procedia Computer Science 54:595
Veerakumar T, Esakkirajan S, Vennila I (2012) . Int J Comput Appl 39(12):29
Wang Z, Bovik AC, Sheikh HR, Simoncelli EP (2004) . IEEE Trans Image Process 13(4):600
Acknowledgements
The author would like to thank Mr. Ankur Bansal, Northeastern University, Bostan, USA for providing the MATLAB software support. The author would also like to thank the editor and the anonymous referees for their valuable comments and suggestions that improved the clarity and quality of this manuscript.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Goel, N. Modified decision based two-phase unsymmetrical trimmed/winsorized mean filter for removal of very high density salt and pepper noise from images and videos. Multimed Tools Appl 81, 32953–32979 (2022). https://doi.org/10.1007/s11042-022-12876-2
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-022-12876-2