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

Table 3

The statistical comparison (Gaussian noise ) between the proposed method and the other methods.

Methods	Proposed	K-SVD	BLS-GSM	BayesShrink	SureShrink
PSNR	31.90 ± 1.55	31.23 ± 2.12	31.10 ± 1.53	29.71 ± 1.03	27.62 ± 0.96
(mean ± SD^1)
values	—	0.038	<0.01	<0.01	<0.01
(proposed versus #^2)

SD: standard deviation of the PSNR series; ²#: one of the other four methods (i.e., SureShrink, BayesShrink, BLS-GSM, and K-SVD).