2015Unpublished venueRequires access

A new adaptive switching approach for impulse noise removal from color images

Kh. Manglem Singh, Keisham Pritamdas

Open publisher page 1 citations

Abstract

A new adaptive switching approach is presented in which the detection of noise is based on entropy of the pixels. Then the detected noisy pixels are replaced with the output of a fuzzy weighted filter. With a little increase in computational complexity over the basic Vector median filter and its varieties, this technique works well both in lower and higher noise ratios. Simulation results show that this method outperforms many other existing nonlinear filters in terms of noise reduction and fine details preservation. This technique can also be extended for images corrupted with Gaussian noise and mixed Gaussian and Impulse noise.

About this research paper

What this paper is about

A new adaptive switching approach is presented in which the detection of noise is based on entropy of the pixels. Then the detected noisy pixels are replaced with the output of a fuzzy weighted filter. With a little increase in computational complexity over the basic Vector median filter and its varieties, this technique works well both in lower and higher noise ratios. Simulation results show that this method outperforms many other existing nonlinear filters in terms of noise reduction and fine details preservation. This technique can also be extended for images corrupted with Gaussian noise and mixed Gaussian and Impulse noise.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

A new adaptive switching approach is presented in which the detection of noise is based on entropy of the pixels. Then the detected noisy pixels are replaced with the output of a fuzzy weighted filter. With a little increase in computational complexity over the basic Vector median filter and its varieties, this technique works well both in lower and higher noise ratios. Simulation results show that this method outperforms many other existing nonlinear filters in terms of noise reduction and fine details preservation. This technique can also be extended for images corrupted with Gaussian noise and mixed Gaussian and Impulse noise.

Key concepts: Impulse noise, Salt-and-pepper noise, Gaussian noise, Value noise, Median filter, Gradient noise, Computer science, Noise reduction

Related papers

Back to paper searchBrowse research topicsOriginal source
A new adaptive switching approach for impulse noise removal from color images — Research Paper | ScholarLens