Improved Adaptive Median Filter Algorithm for Removing Impulse Noise from Grayscale Images


School of Mechanical, Electrical and Information Engineering, Shandong University, Weihai, China


Digital image is often degraded by many kinds of noise during the process of acquisition and transmission. To make subsequent processing more convenient, it is necessary to decrease the effect of noise. There are many kinds of noises in image, which mainly include salt and pepper noise and Gaussian noise. This paper focuses on median filters to remove the salt and pepper noise. After summarizing the main disadvantages of the conventional median filters, this paper proposes a new kind of median filter algorithm based on the detection of impulse noise points. The performance of the proposed algorithm is compared with the conventional standard median filter (SMF), extremum median filter (EMF), and adaptive median filter (AMF). Experimental results under various noise intensities show that the proposed method has better denoising performance and detail preservation compared with the other denoising methods.


1.     Teoh, S.H. and Ibrahim, H., "Median filtering frameworks for reducing impulse noise from grayscale digital images: A literature survey", International Journal of Future Computer and Communication,  Vol. 1, No. 4, (2012), 323-330.
2.     Hwang, H. and Haddad, R.A., "Adaptive median filters: New algorithms and results", IEEE Transactions on Image Processing,  Vol. 4, No. 4, (1995), 499-502.
3.     Juneja, M. and Mohana, R., "An improved adaptive median filtering method for impulse noise detection", International Journal of Recent Trends in Engineering,  Vol. 1, No. 1, (2009), 274-278.
4.     Singh, K.K., Mehrotra, A., Nigam, M.J. and Pal, K., "A novel edge preserving filter for impulse noise removal", in Multimedia, Signal Processing and Communication Technologies (IMPACT), 2011 International Conference on, IEEE., (2011), 103-106.
5.     Wu, J. and Tang, C., "PDE-based random-valued impulse noise removal based on new class of controlling functions", IEEE Transactions on Image Processing,  Vol. 20, No. 9, (2011), 2428-2438.
6.     Tian, H., Cai, H. and Lai, J., "A novel diffusion system for impulse noise removal based on a robust diffusion tensor", Neurocomputing,  Vol. 133, (2014), 222-230.
7.     Fabijanska, A. and Sankowski, D., "Noise adaptive switching median-based filter for impulse noise removal from extremely corrupted images", IET Image Processing,  Vol. 5, No. 5, (2011), 472-480.
8.     Atulkar, M., Zadgaonkar, A.S. and Kumar, S., "Impulse noise removal technique based on fuzzy logic ", Transactions on Machine Learning and Artificial Intelligence,  Vol. 2, No. 6, (2014), 99-105.
9.     Tukey, J.W., "Exploratory data analysis", Journal of the American Statistical Association, Vol. 26, No. 2, (1977), 163-182.
10.   Dong, J. and Zhang, J., "A simple algorithm for removing salt and pepper noise from gray-scale image", Computer Engineering and Applications,  Vol. 39, (2003), 27-28, 65.
11.   Liu, J.-Y. and Fei, R.-C., "An improved median filtering algorithm for salt and pepper noise", Journal of Liaoning Institute of Science and Technology,  Vol. 10, No. 4, (2008), 24-26.
12.   Xing, C.-J., Wang, S.-J., Deng, H.-J. and Luo, Y.-J., "A new filtering algorithm based on extremum and median value", Journal of Image and Graphics,  Vol. 6, No. 6, (2001), 533-536.
13.   Wang, J.-Y., Zhou, X.-G. and Liao, Q.-Z., "A mixed noise filter based on median-fuzzy technology", Journal of Electronics and Information Technology,  Vol. 28, No. 5, (2006), 901-904.