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

Table 4

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

	1	6	12	20	32
Couple
Noisy	15.0949	22.7065	25.7106	27.9160	29.9535
SureShrink	20.0332	25.8413	27.0039	27.6973	28.2485
BayesShrink	19.3898	25.5683	27.0603	28.9848	30.8208
BLS-GSM	21.9560	26.4502	29.4929	30.5549	32.4309
K-SVD	21.3902	26.2782	29.6724	31.1942	33.2195
Proposed	22.4516	27.0145	30.2409	31.6080	33.9895
Cameraman
Noisy	14.8480	22.4357	25.4539	27.6675	29.7107
SureShrink	20.2833	25.4857	27.3508	29.1526	30.8918
BayesShrink	19.6823	25.1516	27.9765	30.0169	31.3640
BLS-GSM	21.6407	27.0976	28.7225	31.49293	33.1347
K-SVD	21.2120	26.6703	28.6297	32.3785	34.6483
Proposed	22.2975	27.5493	29.7168	32.6638	34.2609
Lena
Noisy	14.8853	22.4857	25.4966	27.7053	29.7469
SureShrink	20.4569	27.5477	28.6355	29.2568	29.7915
BayesShrink	21.8229	29.8639	31.8440	33.2049	34.4592
BLS-GSM	22.8901	30.5525	32.6039	33.5162	35.9007
K-SVD	22.2918	30.1317	32.1205	33.0830	35.3714
Proposed	23.1152	30.9638	32.9499	33.2915	35.7910