A soft weighted median filter for removing general purpose image noise

Siprianus Septian Manek, Handayani Tjandrasa

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Citations (Scopus)

Abstract

In this paper, we propose a new simple method for removing all types of image noise using Soft Weighted Median Filter (SWMF). For each pixel in image, firstly we determine the window to search neighboring pixels. Then all the pixel values in the window are sorted and divided into three parts, the pixel value in part two are considered noise-free pixels, while the pixel values in part one and part three are considered noise candidates. Finally, the noise candidate is replaced by the weighted median filter value of the current window, while the noise-free pixels are left unchanged. The experimental results on ten images with different characteristics indicate that our proposed method is successful for all noise types with an average 24.26 db PSNR value and high restoration quality especially for gaussian, poisson, speckle, and localvar noise.

Original languageEnglish
Title of host publicationProceedings of the 11th International Conference on Information and Communication Technology and System, ICTS 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages25-30
Number of pages6
ISBN (Electronic)9781538628256
DOIs
Publication statusPublished - 19 Jan 2018
Event11th International Conference on Information and Communication Technology and System, ICTS 2017 - Surabaya, Indonesia
Duration: 31 Oct 201731 Oct 2017

Publication series

NameProceedings of the 11th International Conference on Information and Communication Technology and System, ICTS 2017
Volume2018-January

Conference

Conference11th International Conference on Information and Communication Technology and System, ICTS 2017
Country/TerritoryIndonesia
CitySurabaya
Period31/10/1731/10/17

Keywords

  • image
  • median filter
  • noise filtering

Fingerprint

Dive into the research topics of 'A soft weighted median filter for removing general purpose image noise'. Together they form a unique fingerprint.

Cite this