< 30% salt-and-pepper noise (Chan's filter) >
Image set: original | corrupted | adamed05 | 4bank | chan
Noise intensity:  5 | 10 | 15 | 20 | 25 | 30 | 35 | 40 | 45 | 50 | 55 | 60 | 65 | 70 | 75 
Files for download: impnoise_030.zip    impnoise_030_seq.zip
PSNR statistics: min: 29.59 dB (100098.png) max: 41.86 dB (299091.png) avg: 34.54 dB
Filter description: Results were obtained by AM-IEPR method based on salt-and-pepper removal proposed by Prof. Chan (link).
Description of improved AM-IEPR method is available here

30% salt-and-pepper noise, 100098.png, PSNR: 29.59 dB

30% salt-and-pepper noise, 105053.png, PSNR: 40.53 dB

30% salt-and-pepper noise, 106020.png, PSNR: 33.92 dB

30% salt-and-pepper noise, 108073.png, PSNR: 32.67 dB

30% salt-and-pepper noise, 118020.png, PSNR: 31.77 dB

30% salt-and-pepper noise, 118035.png, PSNR: 36.22 dB

30% salt-and-pepper noise, 12003.png, PSNR: 34.5 dB

30% salt-and-pepper noise, 124084.png, PSNR: 36.36 dB

30% salt-and-pepper noise, 145014.png, PSNR: 30.64 dB

30% salt-and-pepper noise, 159045.png, PSNR: 32.23 dB

30% salt-and-pepper noise, 161062.png, PSNR: 36.92 dB

30% salt-and-pepper noise, 176035.png, PSNR: 34.95 dB

30% salt-and-pepper noise, 187003.png, PSNR: 31.81 dB

30% salt-and-pepper noise, 189003.png, PSNR: 31.63 dB

30% salt-and-pepper noise, 189011.png, PSNR: 38.59 dB

30% salt-and-pepper noise, 209070.png, PSNR: 36.92 dB

30% salt-and-pepper noise, 253036.png, PSNR: 40.11 dB

30% salt-and-pepper noise, 299091.png, PSNR: 41.86 dB

30% salt-and-pepper noise, 317080.png, PSNR: 34.4 dB

30% salt-and-pepper noise, 385028.png, PSNR: 32.35 dB

30% salt-and-pepper noise, 56028.png, PSNR: 31.19 dB

30% salt-and-pepper noise, 65010.png, PSNR: 33.1 dB

30% salt-and-pepper noise, 67079.png, PSNR: 33.38 dB

30% salt-and-pepper noise, 97017.png, PSNR: 33.39 dB
Result were obtained with the assistance of Dong Yiqiu and Prof. Raymond Chan

The Berkley Segmentation Dataset and Benchmark