A Maximum-Minimum Filtering Algorithm Base on Threshold

Abstract:

In allusion to the contradiction between salt-and-pepper noise attenuation and image detail-preserving, a maximum-minimum filtering algorithm base on threshold is proposed in this paper. The gray values of the pixels in the filtering windows are given firstly. If the test pixel value is the extreme value of the filtering windows, this pixel is a suspicious noise point. And then sort the absolute value of the difference between the suspicious noise point value and pixels gray value in its neighborhood. Judge the pixel is a real noise point according to the average value of the middle two values in the difference sequence. The original gray value of noise point is determined by the maximum-minimum relationship between the pixel values of the neighborhood. If the pixel is not a noise point, its value is kept. The simulation results show that the proposed algorithm has greatly impact on signal to noise ratio and mean square error which retains more image details while making sure to remove noise.

• Periodical: Applied Mechanics and Materials (Volumes 236-237)
• Pages: 1101-1105
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.236-237.1101
• Online since: November 2012
• Authors: Hong She Dang, Li Na Tian, Fang Zhang
• Keywords: Maximum-Minimum Method, Median Filter, Noise Detection, Salt and Pepper Noise

References

[1] Vladimir Crnojevic´, Nemanja I. Petrovic´. Impulse Noise Filtering Using Rubust Pixel-Wise E-Estimate of Variance, J. EURASIP Journal on Advances in Signal Processing, 2010, 10: 1-10.

DOI: 10.1155/2010/830702

[2] Xiuling Li, Jinsong He. Weighted Median Filter Based on Rotation Invariant Pixel Correlations, J. Journal of Image and Graphics, 1(2010) 56-62.

[3] S. Deivalakshmi, S. Sarath, P. Palanisamy. Detection and Removal of Salt and Pepper Noise in Images by Improved Median Filter, C. IEEE Transactions on Recent Advances in Intelligent Computational Systems, 2011, 363-368.

DOI: 10.1109/raics.2011.6069335

[4] Zhengwang Ruan, Jianzhou Zhang, Liang Zhang. L1 Based Local Edge-Preserving Salt-and-Pepper Denoising, J. Journal of Image and Graphics, 15(2010) 867-872.

[5] Baoguo Wei. Improved adaptive median filtering, J. Computer Applications, 7(2008) 1732-1734.

DOI: 10.3724/sp.j.1087.2008.01732

[6] Jyh Yeong Chang, Jialin Chen. Classifier Augmented Median Filters for Image Restoration, J. IEEE Transactions on Instrumentation and Measurement, 53(2004) 351-356.

DOI: 10.1109/tim.2003.822716

[7] Cangju Xing, Shoujue Wang, Haojiang Deng, eta1．A new filtering algorithm based on extremum and median value, J. Journal of Image and Graphics, 6(2001) 533-536.