Please use this identifier to cite or link to this item:
https://dspace.ncfu.ru/handle/20.500.12258/23510
Title: | New method for detecting and removing random-valued impulse noise from images |
Authors: | Lyakhov, P. A. Ляхов, П. А. Orazaev, A. R. Оразаев, А. Р. |
Keywords: | Image processing;Median filter;Impulse noise;Adaptive filter |
Issue Date: | 2023 |
Citation: | Lyakhov, P.A., Orazaev, A.R. New method for detecting and removing random-valued impulse noise from images // Computer Optics. - 2023. - 47(2), pp. 262-271. - DOI: 10.18287/2412-6179-CO-1145. |
Series/Report no.: | Computer Optics |
Abstract: | The paper proposes a method for detecting and removing impulse noise in images, which determines the similarity between pixels by distance and the difference in brightness values in the local detector window. An impulse noise model is considered, where distorted pixels take random values and also randomly appear in the image. Pixels identified as pulses are recovered with an adaptive median filter. The impulses are determined in the detector window, whose size is calculated in the Euclidean metric and increases with increasing noise intensity in the image. In the experimental part, we discuss visual differences between familiar methods and the one proposed herein on three images for three different impulse noise intensities. In the approximation of image fragments, it is seen that the proposed method copes with the task in the best way, which is also confirmed by numerical estimates of the quality of image reconstruction from impulse noise based on the peak signal-to-noise ratio and the structural similarity index. The proposed method can be used when solving problems of cleaning images under conditions of distorting impulses and for eliminating distortions caused by adverse weather effects, such as raindrops and snow. |
URI: | http://hdl.handle.net/20.500.12258/23510 |
Appears in Collections: | Статьи, проиндексированные в SCOPUS, WOS |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
scopusresults 2554 .pdf Restricted Access | 111.77 kB | Adobe PDF | View/Open | |
WoS 1611 .pdf Restricted Access | 1.95 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.