An efficient adaptive median filtering approach for the removal of impulse noise
Varnita Khare, Shaila Chugh
Abstract
Varnita Khare, Shaila Chugh
Abstract
This paper proposes an efficient and fast approach for filtering the salt and pepper and random valued impulse noise. The algorithm first implies the simple statistical operations for the identification of noisy pixel then after proposes the noisy pixel by an adaptive median filter based approach. The simple architecture and the utilization of computationally efficient mathematical operations for the processing makes the algorithm very fast. The proposed algorithm works well for low density as well as very high density salt and pepper noise and it also filters the random valued impulse noise efficiently. Finally the simulation results of denoising the images corrupted by salt and pepper and random valued impulse noise with the noise density up to 99 % has been presented. The results of proposed method are also compared with the other existing method, and the comparison shows the performance improvement of our method by acceptable margin.
OpenAlex reports 9 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.
This paper proposes an efficient and fast approach for filtering the salt and pepper and random valued impulse noise. The algorithm first implies the simple statistical operations for the identification of noisy pixel then after proposes the noisy pixel by an adaptive median filter based approach. The simple architecture and the utilization of computationally efficient mathematical operations for the processing makes the algorithm very fast. The proposed algorithm works well for low density as well as very high density salt and pepper noise and it also filters the random valued impulse noise efficiently. Finally the simulation results of denoising the images corrupted by salt and pepper and random valued impulse noise with the noise density up to 99 % has been presented. The results of proposed method are also compared with the other existing method, and the comparison shows the performance improvement of our method by acceptable margin.
Key concepts: Salt-and-pepper noise, Impulse noise, Median filter, Computer science, Noise reduction, Gradient noise, Noise (video), Gaussian noise