Improvement of nonlinear diffusion equation using relaxed geometric mean filter for low PSNR images

Ehsan Nadernejad

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

A new method to improve the performance of low PSNR image denoising is presented. The proposed scheme estimates edge gradient from an image that is regularised with a relaxed geometric mean filter. The proposed method consists of two stages; the first stage consists of a second order nonlinear anisotropic diffusion equation with new neighboring structure and the second is a relaxed geometric mean filter, which processes the output of nonlinear anisotropic diffusion equation. The proposed algorithm enjoys the benefit of both nonlinear PDE and relaxed geometric mean filter. In addition, the algorithm will not introduce any artefacts, and preserves image details, sharp corners, curved structures and thin lines. Comparison of the results obtained by the proposed method, with those of other methods, shows that a noticeable improvement in the quality of the denoised images, that were evaluated subjectively and quantitatively, is produced.
Original languageEnglish
JournalElectronics Letters
Volume49
Issue number7
Pages (from-to)457-458
ISSN0013-5194
DOIs
Publication statusPublished - 2013

Fingerprint

Dive into the research topics of 'Improvement of nonlinear diffusion equation using relaxed geometric mean filter for low PSNR images'. Together they form a unique fingerprint.

Cite this