Image De-noising Performance of Rank filters
D. Regan, Shenbagalakshmi Gunasekaran · International Journal of Advanced Research in Computer Science · 2011
In image processing, the rank filters are used to yield better noise removal ability for salt & pepper type noise. But they won’t reproduce the images exactly for high noise density conditions. During image processing, varying noise density surely affects the de-noising performance of the processing system. So, The Noise removal ability of system should be improved. Proper rank filter selection would be important for this. Otherwise we may lose our precise information. In this work, three rank filter algorithms are compared to estimate their performance in high noise density. We can infer the filters response from the results shown. Better Filtering selection can be achieved using PSNR values presented from original and de-noised images. Keywords: Image De-noising, Median filter, Adaptive median filter, Hybrid median filter, PSNR, Salt & pepper noise.