Adaptive median-weighted mean hybrid filter
Xuming Zhang, Binshi Xu
Abstract
Xuming Zhang, Binshi Xu
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.
OpenAlex reports 13 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
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)