Image Defogging Algorithm Based on Sparse Representation

Table 3

Objective evaluation of image quality.

Defog image	Dehazing algorithm				SSIM	PSNR
First image	Reference [5]	0.0058	1.7231	1.5732	0.8765	20.3072
	Reference [6]	0.2104	1.3137	0.5024	0.7214	14.4050
	Reference [10]	0.0354	1.5405	0.3406	0.8452	18.0232
	Our algorithm	0.0137	1.4738	0.0297	0.9685	32.0796
Second image	Reference [5]	0.6265	2.0012	1.5658	0.7043	16.1828
	Reference [6]	0.7527	1.6474	0.1283	0.6367	11.7532
	Reference [10]	0.5980	2.0704	0.1386	0.7789	16.8623
	Our algorithm	0.6936	2.2598	0.0004	0.8906	20.1899
Third image	Reference [5]	2.5911	2.2208	1.5100	0.5485	13.8240
	Reference [6]	2.7239	2.1715	0.0516	0.6389	13.9329
	Reference [10]	1.9947	2.6608	0.0150	0.8578	20.6405
	Our algorithm	1.7333	2.5235	0.0003	0.5248	18.7909
Fourth image	Reference [5]	1.7611	2.4218	1.7546	0.6269	13.8187
	Reference [6]	1.9408	1.6657	0.4637	0.5522	10.9472
	Reference [10]	1.9161	1.2714	1.7076	0.4173	11.2976
	Our algorithm	0.9178	1.5066	0.0005	0.8669	19.5566
Fifth image	Reference [5]	0.9754	2.0911	1.5723	0.6907	15.3297
	Reference [6]	1.1465	1.8062	0.2679	0.6141	11.7734
	Reference [10]	0.9276	2.3240	0.1936	0.8027	17.2055
	Our algorithm	1.5758	1.5387	0.0005	0.7298	19.5380
Mean of five images	Reference [5]	1.1920	2.0916	1.5951	0.6894	15.8925
	Reference [6]	1.3548	1.7209	0.2828	0.6327	12.5623
	Reference [10]	1.0944	1.9734	0.47911	0.7404	16.8058
	Our algorithm	0.9868	1.8605	0.0063	0.7961	22.0310