Variable Splitting Based Method for Image Restoration with Impulse Plus Gaussian Noise

Table 4

Computational results on random-valued noise plus Gaussian noise removal by solving model (7).

Blur		HNW09	ADM2	ADM3	ADMGB
Gaussian	20%	30/40.95/19.08	24/43.78/19.29	42/6.28/19.28	43/8.77/19.25
	30%	31/41.80/18.91	24/41.80/19.01	43/6.28/19.14	45/9.64/19.12
	40%	31/39.66/18.50	25/42.42/18.61	43/6.33/18.54	45/9.17/18.77
	50%	33/41.25/17.40	27/47.44/17.22	47/6.89/17.28	49/10.03/17.26
Out-of-focus	20%	27/35.69/19.46	21/38.78/19.77	38/5.78/19.73	40/8.33/19.73
	30%	27/36.95/19.20	22/40.64/19.45	40/5.91/19.40	42/8.80/19.50
	40%	30/40.58/18.48	23/42.22/18.95	41/6.19/18.93	45/9.30/18.66
	50%	39/48.56/17.05	25/44.56/17.15	45/6.63/17.67	49/10.19/17.20