Research Article

A Novel Directionlet-Based Image Denoising Method Using MMSE Estimator and Laplacian Mixture Distribution

Table 2

PSNR comparison (Gaussian noise) of the proposed method with other methods (in dB).

51020305075100

Boat
Noisy34.117228.139822.189018.731014.604511.748610.1119
SureShrink32.809630.667027.885426.004223.173820.696419.3363
BayesShrink35.700532.548129.158127.219724.961622.944121.4948
BLS-GSM36.732933.483130.299428.436026.178224.564823.5208
K-SVD37.216033.591830.336828.411725.896224.050422.7441
Proposed37.673334.051730.778929.358426.603524.838923.6182

Peppers
Noisy34.134928.172822.222518.775714.737411.861310.2132
SureShrink33.482731.783129.125626.919324.232121.182519.6603
BayesShrink36.145033.343230.100028.165025.969423.244921.7568
BLS-GSM37.113933.943630.752928.667526.084224.259122.5839
K-SVD37.721534.262830.818929.049326.145323.695721.8496
Proposed37.164134.215631.572829.872227.465825.535624.1813

Couple
Noisy34.146828.150122.158018.725314.557011.717310.1078
SureShrink32.572928.980026.630624.986622.375920.249319.3832
BayesShrink36.252732.271428.609426.829424.692522.823321.5051
BLS-GSM36.956432.465429.384027.210124.735723.051521.5715
K-SVD37.268533.425829.672527.547525.119923.492122.4298
Proposed37.063133.018629.970527.783525.505923.900322.8189

Cameraman
Noisy31.489327.081221.926718.828014.984612.298410.6453
SureShrink33.652331.030027.994125.897822.809919.898518.7697
BayesShrink38.231834.186230.461128.193825.058422.521020.4949
BLS-GSM40.356436.545632.982029.817127.634725.065122.7713
K-SVD40.921737.125733.206630.947428.456425.956724.0799
Proposed40.831537.249333.467231.264528.802626.749425.2453

Lena
Noisy30.018926.283721.396918.350014.594011.937810.3692
SureShrink31.683729.711327.519125.983823.635221.236620.1037
BayesShrink36.830733.417530.219628.590526.298324.069522.3581
BLS-GSM38.472935.644132.598529.710128.114926.586225.6407
K-SVD38.580635.487432.352230.413327.765325.773124.4709
Proposed38.749335.849333.167231.264528.202626.449425.0453