2004Optical TechniqueRequires access

Adaptive median-weighted mean hybrid filter

Xuming Zhang, Binshi Xu

Open publisher page 13 citations

Abstract

To remove mixed impulse noise and Gaussian noise in digital images, a hybrid filter based on the adaptive median filter and the adaptive weighted mean filter are proposed. The algorithm first divides the image into several parts and separates the two kinds of noise in each part by means of noise detection, then applies the adaptive median filter and the adaptive weighted mean filter to removing impulse noise and Gaussian noise, respectively. Results of computer simulation experiments on this new filter show that it has better filtering performance than the three hybrid filters mentioned, and it can suppress impulse noise and Gaussian noise effectively and preserve fine details very well, which provides an effective means for removing mixed noise in digital images.

About this research paper

What this paper is about

To remove mixed impulse noise and Gaussian noise in digital images, a hybrid filter based on the adaptive median filter and the adaptive weighted mean filter are proposed. The algorithm first divides the image into several parts and separates the two kinds of noise in each part by means of noise detection, then applies the adaptive median filter and the adaptive weighted mean filter to removing impulse noise and Gaussian noise, respectively. Results of computer simulation experiments on this new filter show that it has better filtering performance than the three hybrid filters mentioned, and it can suppress impulse noise and Gaussian noise effectively and preserve fine details very well, which provides an effective means for removing mixed noise in digital images.

Why it matters

OpenAlex reports 13 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

To remove mixed impulse noise and Gaussian noise in digital images, a hybrid filter based on the adaptive median filter and the adaptive weighted mean filter are proposed. The algorithm first divides the image into several parts and separates the two kinds of noise in each part by means of noise detection, then applies the adaptive median filter and the adaptive weighted mean filter to removing impulse noise and Gaussian noise, respectively. Results of computer simulation experiments on this new filter show that it has better filtering performance than the three hybrid filters mentioned, and it can suppress impulse noise and Gaussian noise effectively and preserve fine details very well, which provides an effective means for removing mixed noise in digital images.

Key concepts: Median filter, Salt-and-pepper noise, Gaussian noise, Impulse noise, Adaptive filter, Computer science, Kernel adaptive filter, Noise (video)

Related papers

Back to paper searchBrowse research topicsOriginal source
Adaptive median-weighted mean hybrid filter — Research Paper | ScholarLens