A scale-based forward-and-backward diffusion process for adaptive image enhancement and denoising

Wang, Yi; Niu, Ruiqing; Zhang, Liangpei; Wu, Ke; Sahli, Hichem

doi:10.1186/1687-6180-2011-22

EURASIP Journal on Advances in Signal Processing

Table 3 PSNR of the six diffusion algorithms for the noisy testing images of Peppers, Lena, Cameraman and Boat with respect to different noise variances.

From: A scale-based forward-and-backward diffusion process for adaptive image enhancement and denoising

Scheme	Image	Noise variance (σ²)
		100		225		400		625		900
		PSNR	UIQI	PSNR	UIQI	PSNR	UIQI	PSNR	UIQI	PSNR	UIQI
CAD	Peppers	32.93	0.5917	31.90	0.5681	30.81	0.5367	29.81	0.5040	28.93	0.4737
	Lena	33.48	0.6518	31.16	0.6118	31.08	0.5733	30.06	0.5339	29.12	0.4961
	Cameraman	34.55	0.5819	32.89	0.5138	31.43	0.4588	30.06	0.4156	28.81	0.3806
	Boat	30.87	0.6252	30.03	0.6048	29.18	0.5816	28.31	0.5507	27.55	0.5252
WAD	Peppers	32.57	0.5771	31.60	0.5553	30.61	0.5287	29.67	0.5001	28.87	0.4719
	Lena	32.98	0.6345	31.84	0.6036	30.80	0.5667	29.87	0.5309	29.00	0.4959
	Cameraman	33.96	0.5619	32.51	0.4984	31.13	0.4487	29.84	0.4072	28.67	0.3722
	Boat	30.55	0.6022	29.73	0.5814	28.88	0.5579	28.09	0.5318	25.37	0.5078
RAD	Peppers	31.44	0.6165	28.27	0.4995	25.82	0.4118	23.95	0.3496	22.50	0.3042
	Lena	31.91	0.6174	28.36	0.4931	25.88	0.4095	23.98	0.3525	22.46	0.3085
	Cameraman	32.60	0.4944	28.81	0.3868	26.21	0.3278	24.23	0.2854	22.61	0.2538
	Boat	31.46	0.7036	28.33	0.6037	25.87	0.5164	23.98	0.4460	22.42	0.3927
MAD	Peppers	32.66	0.6025	30.84	0.5538	28.97	0.4930	27.28	0.4373	25.95	0.3919
	Lena	33.32	0.6583	31.19	0.5886	29.19	0.5137	27.54	0.4552	26.09	0.4046
	Cameraman	34.15	0.5809	31.63	0.4773	29.24	0.3990	27.33	0.3453	25.77	0.3071
	Boat	31.25	0.6475	29.74	0.6103	28.14	0.5599	26.63	0.5050	25.34	0.4578
EED	Peppers	33.04	0.6130	31.62	0.5754	30.15	0.5274	28.88	0.4832	27.76	0.4447
	Lena	33.85	0.6702	32.11	0.6128	30.60	0.5608	29.23	0.5104	28.01	0.4656
	Cameraman	34.77	0.5952	32.72	0.5088	30.87	0.4508	29.28	0.4045	27.90	0.3690
	Boat	31.28	0.6655	30.26	0.6348	29.14	0.6018	28.08	0.5613	27.07	0.5282
SFABD (without EODDS)	Peppers	32.94	0.5977	31.82	0.5662	30.80	0.5366	29.90	0.5071	28.99	0.4743
	Lena	33.76	0.6612	32.13	0.6135	30.95	0.5725	30.01	0.5338	29.24	0.5032
	Cameraman	35.06	0.5947	33.01	0.5159	31.41	0.4572	30.17	0.4205	29.05	0.3858
	Boat	31.44	0.6496	29.92	0.6043	28.79	0.5681	27.93	0.5363	27.20	0.5093
SFABD (with EODDS)	Peppers	33.33	0.6210	32.03	0.5801	30.90	0.5407	30.08	0.5109	29.35	0.4789
	Lena	34.24	0.6763	32.51	0.6195	31.21	0.5737	30.28	0.5378	29.67	0.4957
	Cameraman	35.64	0.6006	33.58	0.5222	31.73	0.4654	30.53	0.4250	29.46	0.3929
	Boat	31.87	0.6805	30.69	0.6363	29.49	0.6090	28.55	0.5684	27.59	0.5330

For the SFABD, the parameters settings are: σ_smooth = 2, σ_edge = 1, δ_smooth = 0.3, δ_detail = 0.6, δ_edge = 0.9

Back to article page