Previous Articles     Next Articles

Study on the image denoising based on the improved median filtering algorithm

  

  1. (1 Jiangsu Polytechnic College of Agriculture and Forestry, Zhenjiang 212400,China; 2 School of Mechanical Engineering, Southeast University , Nanjing 211189, China;3 Nanjing Communications Institute of Technology, Nanjing 211189, China)
  • Online:2015-06-25 Published:2015-06-26

Abstract: In order to remove the high\|density salt and pepper noise, and preserve the details of the original image well, an improved filtering method was proposed based on the median filter and the mean filter in this paper. In the first phase, it was necessary to identify pixels which were likely to be noise. In the second phase, the image was filtered recursively by the small window, salt and pepper pixel in the current window was calculated, then the noise pixel using the median and mean weighted value of the non\|noise pixel were updated. The experimental results showed that the proposed method behaved excellently for all noise ratios, from 5% to 80%. This method has achieved the purpose of the good filtering effect on the subjective quality and objective quality, which was beneficial for the further analysis and interpretation of images.

Key words: median filtering, salt and pepper noise, image denoising